Towards Optimal Diagnosis of Type II Germ Cell Tumors

Towards optimal diagnosis of type II germ cell tumors

Hans Stoop

Towards optimal diagnosis of type II germ cell tumors

Optimale diagnostiek van type II kiemceltumoren

The work presented in this thesis was done at the Department of Pathology, Erasmus Medical Center, Josephine Nefkens Institute, Daniel den Hoed Cancer Center, Rotterdam, The Netherlands

Cover design and document layout: Jackelien Karel-Stoop & Martin Rijlaarsdam

Printed by: Ridderprint BV, Ridderkerk

ISBN: 978-90-5335-482-7

Towards Optimal Diagnosis of Type II Germ Cell Tumors

Optimale diagnostiek van type II kiemceltumoren

Proefschrift

ter verkrijging van de graad van doctor aan de Erasmus Universiteit Rotterdam

op gezag van de rector magnificus prof.dr. H.G. Schmidt

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 30 november 2011 om 9.30 uur

door

Johannes Adrianus Stoop

geboren te Bergen op Zoom

PROMOTIECOMMISSIE:

Promotoren: Prof.dr. L.H.J. Looijenga

Prof.dr. J.W. Oosterhuis

Overige leden: Prof.dr. C.H. Bangma

Prof.dr. S.L.S. Drop

Prof.dr. R.R. de Krijger

Our wisdom comes from our experience, and our experience comes from our foolishness

Sacha Guitry

Voor mijn ouders: Sjaak † en Lisa

Contents

Abbreviations

Chapter 1 13 General Introduction 1.1 Introduction 1.2 Migration of primordial germ cells 1.3 Gonocyte micro-environment 1.4 Differentiation into testis 1.5 Classification of germ cell tumors 1.5.1 Type III germ cell tumors 1.5.2 Type II germ cell tumors 1.5.2.1 Introduction 1.5.2.2 Etiology and identified risk factors 1.5.2.3 Compiling model of type II (testicular) germ cell tumor risk factors 1.5.2.4 Precursor of testicular germ cell tumors 1.6 Concluding remarks References

Chapter 2 27 Aims and outlines of the thesis

Chapter 3 31 Pathobiological implications of the expression of markers of testicular carcinoma in situ by fetal germ cells. Journal of Pathology 203: 849-857, 2004

Chapter 4 49 Germ cell lineage differentiation in non-seminomatous germ cell tumours. Journal of Pathology 208: 395-400, 2006

Chapter 5 61 POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Research 63: 2244-2250, 2003

Chapter 6 79 Diagnostic value of OCT3/4 for pre-invasive and invasive testicular germ cell tumours. Journal of Pathology 206: 242-249, 2005

Chapter 7 95 Non-invasive detection of testicular carcinoma in situ in semen using OCT3/4. European Urology 54: 153-160, 2008

Chapter 8 113 A 40-year-old woman with a progressive periventricular white matter lesion. Brain Pathology 18: 103-104/142, 2008

Chapter 9 119 Differential expression of SOX17 and SOX2 in germ cells and stem cells has biological and clinical implications. Journal of Pathology 215: 21-30, 2008

Chapter 10 137 Stem cell factor as a novel diagnostic marker for early malignant germ cells. Journal of Pathology 216: 43-54, 2008

Chapter 11 157 Diagnosis of testicular carcinoma in situ (intratubular and micro-invasive) seminoma and embryonal carcinoma using direct enzymatic alkaline phosphatase reactivity on frozen histological sections. Histopathology 58: 440-446, 2011

Chapter 12 169 Global DNA methylation in fetal human germ cells and germ cell tumours: association with differentiation and cisplatin resistance. Journal of Pathology 221: 433-442, 2010

Chapter 13 187 General Discussion 13.1 Introduction 13.2 Normal testicular development and protein expression profile 13.3 From pluripotency to totipotency of (testicular) germ cell tumors 13.4 OCT3/4 as ultimate marker for diagnosis and non-invasive detection 13.5 Differential diagnosis of seminomatous tumors and embryonal carcinoma 13.6 SCF as diagnostic marker and early pathogenetic mechanism 13.7 Direct frozen tissue diagnosis of CIS and invasive testicular germ cell tumors 13.8 Epigenetics 13.9 Concluding remarks References

Chapter 14 201 Summary/Samenvatting

Dankwoord 211

Curriculum Vitae 213

List of publications 214

Abbreviations

AFP: a-fetoprotein ChC: choriocarcinoma CIS: carcinoma in situ c-KIT: stem cell factor receptor (CD117) dAP: direct alkaline phosphatase (staining) DSD: disorders of sex development EC: embryonal carcinoma FFPE: formalin fixed paraffin embedded GB: gonadoblastoma GBY: critical region of the gonadoblastoma locus on Y chromosome GCAP: germ cell alkaline phosphatase GCL: germ cell lineage H&E: haematoxylin and eosin (staining) hCG: human chorionic gonadotrophin IGCNU: intratubular germ cell neoplasia unclassified IHC: immunohistochemistry KITLG: c-KIT ligand (SCF) LDH: lactate dehydrogenase PGC(s): primordial germ cell(s) PLAP: placenta like alkaline phosphatase pSg: pre-spermatogonium Sc: spermatocyte SCF: stem cell factor (KITLG) SCO: sertoli cell only SE: seminoma Sp: spermatogonium Sz: spermatozoa TDS: testicular dysgenesis syndrome TE: teratoma TGCT(s): testicular germ cell tumors TIN: testicular intratubular neoplasia TM: testis microcalcifications TMA: tissue micro array TNAP: tissue non-specific alkaline phosphatase TSAP: tissue specific alkaline phosphatase TSPY: testis specific protein Y encoded YST: yolk sac tumor

A:IM>KØ

8A8E4?AGEB7H6G

CHAPTER 1

1.1 Introduction Primordial Germ Cells (PGCs) are the germ line stem cells in mammals that give rise to gametes at later life. These final mature germ cells, oocytes in females and spermatids in males, are designed to transfer genetic information to the next generation. Therefore they are specifically and optimally equipped, and undergo defined steps of maturation and specialization during their development. Much research on the fundamental aspects of germ cell development is done in vivo or ex vivo on animals, including Drosophila, zebrafish 1, and mouse 2-6. These observations are extrapolated to the human situation, because of the limitations to study early human embryogenesis. However, this has restrictions, because of possible differences between mice and humans, including and specifically the germ cell lineage at early (embryonic) age 7. Therefore direct investigation of human germ cell development is crucial to understand the various processes involved, and to prevent wrong interpretation due to the existence of interspecies variation(s). This is especially of relevance in the context of understanding the earliest events in the pathogenesis of human germ cell tumors (GCTs), for which significant differences exist between humans and animals 8. In the next paragraphs a selection of relevant observations will be presented, informative for better understanding the pathogenesis of human GCTs. Because of this defined goal, the information given will be no review of the whole existing literature on this topic, for which other papers are available 8-10.

1.2 Migration of primordial germ cells In the third week of human embryonic development PGCs start to migrate by active amoeboid movement under response of mitogenic and survival factors from their origin in the posterior wall of the yolk sac along the hindgut, through the dorsal mesentery into the genital ridge 1,3,6,11-12. This migration route can be visualized based on the fact that PGCs express a number of specific (embryonic) markers, involved in various biological mechanisms during this time window. Here, only those markers that have a (putative) value in diagnosis of human GCTs, i.e., alkaline phosphatase, OCT3/4, c-KIT, VASA, TSPY and SOX17 13-17 will be referred to. Proper PGC development is dependent on the c-KIT pathway, in which both a survival (i.e., apoptosis suppressing) effect and well as a proliferative effect of the ligand (KITLG or Stem Cell Factor) can be identified 5. In the rest of the introductory chapter the human embryonic gonadal male situation will be described, if not stated otherwise.

14 General Introduction

1.3 Gonocyte micro-environment At the time the PGCs enters the genital ridge (later in the fifth week of development), they are called gonocytes. That moment, their migratory behavior ceases, but they continue to proliferate (the number at that time is approximately 1000 to 2000) 18. At this specific moment of development, no differences between the sexes exist, related to both the germ cells itself as well as their micro-environment. However, early in the sixth week in the male constitution (XY), initiated by expression of the transcription factor SRY (testis determining factor), and subsequently SOX-9, Sertoli cells are formed, resulting in the formation of primitive sex cords, wherein the gonocytes interact with the supportive cells. This will be presented in more detail in the next paragraph.

1.4 Differentiation into testis Under influence of SRY, of which the gene is mapped to the Y chromosome, the first differences between the male and female gonad are formed. In fact, the bipotential gonad is triggered to differentiate into the testicular direction by inducing a subset of somatic cells to differentiate into Sertoli cells, believed to act as the organizing center of the male gonad 19. The presence of germ cells plays no part in this specific process in the male 20. Parallel to the differentiation and proliferation of the Sertoli cells, the gonad increases in size due to increased proliferation and migration of other cells which will give rise to endothelial-, myoepithelial-, and Leydig cells 18,21-23. During the 7th week of gestation the gonocytes become enclosed in the further developed sex cords, now referred to as primitive seminiferous cords, and enter mitotic arrest. At this time, the amount of germ cells is about 100.000 24-26. This situation will remain stable until a few days after birth. The gonocytes present have a high content of glycogen as in the stage of PGC, but still show expression of the other proteins, as mentioned before (alkaline phosphatase, OCT3/4, c-KIT, VASA, TSPY and SOX17) 27-30. The gonocytes and Sertoli cells have initially a non-structured organization, but gradually the germ cells migrate toward the periphery of the further developing seminiferous tubules. In the 13th week the germ cells are predominantly localized at the basal side, and start to loose expression of the embryonic markers mentioned, with the exception of VASA and SOX17. They start to further enhance expression of TSPY (Testis Specific Protein on the Y chromosome) 31. Contact between Sertoli cells and, the now referred to, pre-spermatogonia is required for proper regulation of proliferation and protection against apoptosis 32-34. In fact, the gonocytes and pre-spermatogonia change their expression pattern gradually; proteins related to an embryonic stage are decreasing in level, and finally disappear. These different steps in the maturation of germ cells are schematically illustrated in Figure 1.

15 CHAPTER 1 A

Figure 1. Maturation and migration of germ cells. (A) The different steps in the maturation of germ cells. Expression of the different diagnostic markers alkaline phosphatase (AP), OCT3/4, c-KIT, VASA, SOX17 and TSPY from the early embryonal period which starts when the PGCs originate from the epiblast. Later in the 5th week when the PGCs enter the genital ridge the induced Sertoli cells surround the PGCs which are now called gonocytes (GC) and the seminiferous tubule is formed. The expression pattern changes and the proteins related to the embryonic stage are declining in level and finally disappear. In puberty spermatogenesis starts; (B) Totipotent PGCs migrate to the genital ridge and start to mature. At the time of entering the genital ridge they are called gonocytes (being pluripotent) and subsequently when they are in contact with Sertoli cells and the basal lamina, are referred to pre- spermatogonia (pSg). In puberty the unipotent spermatogonia (Sg) form in succession spermatocytes (Sc) and spermatozoa (Sz).

16 General Introduction

1.5 Classification of germ cell tumors The traditional classification of human GCTs is primarily based on histology. A new classification system, recognized by the World Health Organization (WHO), is based on various clinical, genetic and histological parameters which makes distinction into five relevant entities possible 35-36. Identification of these subtypes will result in an improved clinical diagnosis and prediction of treatment response, as well as more straightforward research. The group of testicular GCTs comprises three of the five GCT entities. The so-called type I GCTs of the testis are predominantly diagnosed in neonates and infants 35-36, and will therefore not be discussed here. The types II and III are the majority (90%) of a diverse group of lesions in the testis at adult age. Other types of testicular tumors include the gonadal stromal tumors, i.e., Leydig cell tumors, Sertoli cell tumors and others 37, also not to be discussed here. The type II and III testicular GCTs can be distinguished from each other based on cell of origin, pathogenesis and clinical behavior. Because of the topic of this thesis, the type III GCTs will be referred to only briefly in the next paragraph, after which the type II GCTs will be discussed throughout.

1.5.1 Type III germ cell tumors Spermatocytic seminomas, here referred to as type III GCTs 36, are a rare variant of testicular GCTs, accounting less than 1% of all cases in adults, with an incidence of about 0,2 per 100.000, without increase in the population 38. According to the WHO classification system, the spermatocytic seminomas are a variant of seminoma 35. However, they are predominantly found at elderly age, show a distinct morphology, and hardly metastasize 39-41. Based on multiple observations, it is currently accepted that type III GCTs originate from a later maturation stage of the germ cell lineage, likely spermatogonia or spermatocyte 42-44. Interestingly, recent observations indicate that gain of chromosome 9 is a recurrent anomaly, for which DMRT1 is a possible candidate, informative as diagnostic marker 45. In addition, other informative markers for diagnosis of type III GCTs are reported 39,42-43.

1.5.2 Type II germ cell tumors

1.5.2.1 Introduction Testicular type II GCTs, here referred to as TGCTs, are histologically and clinically subdivided into the seminomatous and nonseminomatous tumors 35. The nonseminomas can be subdivided into the stem cell component, referred to as embryonal carcinoma, the somatic lineages (teratoma) and the extra-embryonic lineages (yolk sac tumor and choriocarcinoma). During the last 40 years the incidence of TGCTs doubled in Caucasian populations 46, while they are

17 CHAPTER 1 significantly less frequent in Asian and African populations, independent of migration. Clear differences are also identified within Europe, with the highest incidence in Switzerland and Denmark 47. In the Netherlands the incidence of TGCTs is about 6.6 per 100.000 corresponding to approximately 750 newly diagnosed cases each year. The incidence is still growing (4.8 per 100.000 in the period 1994-1998) 38,46,48-50. In spite of this rise there is no increase in death of patients because of this disease. The reason is the availability of effective treatment options, including surgery, irradiation and/or chemotherapy, resulting in a 10 year survival of about 95% 51. In the Netherlands, before 1977 (the pre- cisplatin period), the 10 year survival of a patient with a seminoma was 83%, and with nonseminoma 56% 38,52. However, there are drawbacks related to the recent success leading to survival of 95%, including long term side effects of systemic treatment, like heart- and vascular disease, infertility, metabolic syndrome, and secondary malignancies 53. Therefore, understanding of the earliest steps in the pathogenesis of TGCTs is of importance, because it will allow development of methods for early diagnosis and treatment, limiting treatment-related side effects. In the next paragraph the current knowledge about the etiology of TGCTs will be summarized.

1.5.2.2 Etiology and identified risk factors As indicated, a remarkable geographical variation exists in the (increase of) incidence of TGCTs 47-48,54. It is suggested that TGCT-development is related to a Western lifestyle 55. Environmental exposures during early development have been postulated to play a significant role 47,55-60. This is demonstrated by analysis of migration studies and the dip in the increasing trend during the Second World War. In addition, a number of defined risk factors is identified. These include cryptorchidism, in fact the first condition recognized as being involved with the etiology of this cancer 61-66. Only 2-10% of the patients with a TGCT have a history of undescended testis 67. This indicates that other risk factors must exist. One of these is infertility 68. Within the overall Andrology clinic population, the incidence of TGCT (or the precursor of it, see below) is 1% 69. In addition, hypospadias and a low or high birth weight are identified as risk factors 59,70-71. Moreover, a previous (unilateral) TGCT is a risk factor as well. The cumulative risk to develop a contralateral TGCT ranges from 0.6 % to 8%. 72-74, being higher in patients with a nonseminoma compared to with a seminoma 73,75. Interestingly, a role of the immune system has been suggested 76-78.

Besides these effects, a significant genetic component is identified 79-80. This is demonstrated by the reported increased risk for this type of cancer in fathers and brothers, although this might also be related to a common micro-environment. However, no high penetrance predisposing genes were discovered by various linkage studies in TGCT-families81-83. Recently significant findings were made

18 General Introduction using genome-wide association (GWA) studies. Independent investigations showed the presence of TGCT-predisposing alleles in or near a limited number of genes, i.e., KITLG (SCF), BAK1, SPRY4, TERT, ATF7IP, and DMRT1 81-87. Interestingly, the proteins encoded by these genes are (partly) known to have a role in early germ cell development, including migration, survival, micro-environment establishment and maturation.

1.5.2.3 Compiling model of testicular germ cell tumor risk factors Skakkebæk and coworkers proposed in 2001 existence of a Testicular Dysgenesis Syndrome (TDS), characterized by impaired spermatogenesis, cryptorchidism and hypospadias, associated with possible development of TGCTs 88. It has been hypothesized that TDS is the result of anti-androgen and/or (xeno)estrogen action, possibly via the supportive cell function 89-91. This may act via genetic- or epigenetic mechanisms. For example phthalates, which are widespread used, are classified as endocrine disruptors 92-95. Exposure can lead to demethylation of promoter-associated CpG islands, a crucial regulator of gene expression, both in time and place 96. A high estrogen level exposure during pregnancy has indeed been suggested 97, supported by a recent study 98. Related to this model, another, so far not mentioned risk factor for type II GCTs is of relevance. These anomalies are described by the umbrella term Disorders of Sex Development (DSD), previously referred to as intersex 99. It is characterized by discordance between genetic sex, gonadal sex and phenotypic sex, in which patients showing hypovirilization and/or gonadal dysgenesis are at increased risk 100. This is directly related to the presence of the gonadoblastoma region of the Y chromosome (GBY), including the forementioned TSPY gene 101. Another parameter for prediction of malignant transformation of the germ cell is the anatomical localization of the gonad, as expected based on identification of cryptorchidism as risk factor (see above). Overall, an abdominal gonad has a higher risk for malignant transformation than a scrotally localized gonad for review 102-105. To allow optimal diagnosis of type II GCTs, both in the testis, as well as gonads of patients with DSD, recognition of the precursor lesion of this type of cancer is required. Current knowledge related to this issue will be presented in the next paragraph.

19 CHAPTER 1

1.5.2.4 Precursor of type II (testicular) germ cell tumors Presence of atypical germ cells in the seminiferous tubules adjacent to a TGCT was detected already in 1896 106. They were seen as a product of the cancer or a representation of spread from the malignancy itself 107. However, the association between these cells and progression to a TGCT was officially established by Skakkebæk in 1972 108. This observation was in line with the earlier suggestion that the various histological variants of TGCTs have a common precursor 109. This precursor is referred to as carcinoma in situ (CIS) (the term to be used here), intratubular germ cell neoplasia unclassified (IGCNU) 110-111 and testicular intratubular neoplasia (TIN) 112. Supportive data that a PGC/gonocyte is the origin of CIS is gathered by analysis of morphology, and protein and (m and mi)RNA expression levels, showing similarities to embryonic stem cells 113-116. Based on these observations, as well as the specific risk factors identified (see above), the currently accepted pathogenetic model of type II (T)GCTs is that they originate early in embryogenesis, affecting a PGC/gonocyte, blocked in its process of maturation 117-118. The presence of CIS predicts development of an invasive TGCT 119-120, which indicates that clinical interference is required. The CIS counterpart in a gonad of a DSD patient with a low level of testis development (referred to as “testicularization” 118), is known as gonadoblastoma (GB) 121. In fact, the germ cells in GB show the same characteristics of CIS cells, including their tendency to progression to an invasive type II GCT. A summary of the various relevant items presented, is given in Figure 2.

Figure 2. Pathogenesis of the type II (testicular) germ cell tumor. The arrested embryonal germ cell (pre-CIS cell) in the embryonal gonad, is supposed to start proliferating at puberty. Via the specific intratubular stage CIS/GB develops in an invasive germ cell tumor: seminoma and/or nonseminoma. The embryonal carcinoma, which is the stem cell component of the nonseminoma, can differentiate into a somatic (teratoma) and/or extra-somatic (yolk sac tumor and choriocarcinoma) entity.

20 General Introduction

1.6 Concluding remarks An interplay between environmental - and genetic factors seems to be involved in the earliest pathogenetic events in development of type II (T)GCTs. This knowledge allows identification of risk populations to develop this type of cancer in the earliest stage of development possible, leading to early diagnosis and treatment, thereby preventing progression to a cancer. This set up requires availability of informative (highly sensitive and specific) markers and tools for detection of the earliest changes in the process of malignant transformation of a germ cell. This specifically applied on histological samples obtained by surgical biopsies of gonads at risk for this type of cancer. In addition, in certain situations, a non-invasive diagnostic test might be applicable.

References 1. Molyneaux, K. & Wylie, C. Primordial germ cell migration. Int J Dev Biol 48, 537-544 (2004). 2. Damjanov, I., et al. Immunohistochemical localization of murine stage-specific embryonic antigens in human testicular germ cell tumors. Am J Pathol 108, 225-230 (1982). 3. Fujimoto, T., Yoshinaga, K. & Kono, I. Distribution of fibronectin on the migratory pathway of primordial germ cells in mice. Anat Rec 211, 271-278 (1985). 4. Donovan, P.J., Stott, D., Cairns, L.A., Heasman, J. & Wylie, C.C. Migratory and postmigratory mouse primordial germ cells behave differently in culture. Cell 44, 831-838 (1986). 5. Donovan, P.J., de Miguel, M., Cheng, L. & Resnick, J.L. Primordial germ cells, stem cells and testicular cancer. Apmis 106, 134-141 (1998). 6. Molyneaux, K.A., Stallock, J., Schaible, K. & Wylie, C. Time-lapse analysis of living mouse germ cell migration. Dev Biol 240, 488-498 (2001). 7. De Miguel, M.P., Fuentes-Julian, S. & Alcaina, Y. Pluripotent stem cells: origin, maintenance and induction. Stem Cell Rev 6, 633-649 (2010). 8. Looijenga, L.H., Gillis, A.J., Stoop, H.J., Hersmus, R. & Oosterhuis, J.W. Chromosomes and expression in human testicular germ-cell tumors: insight into their cell of origin and pathogenesis. Ann N Y Acad Sci 1120, 187-214 (2007). 9. Looijenga, L.H. & Oosterhuis, J.W. Pathobiology of testicular germ cell tumors: views and news. Anal Quant Cytol Histol 24, 263-279 (2002). 10. Honecker, F., et al. New insights into the pathology and molecular biology of human germ cell tumors. World J Urol 22, 15-24 (2004). 11. Fujimoto, T., et al. The primordial germ cells in amniotes: their migration in vivo and behaviors in vitro. Prog Clin Biol Res 296, 13-21 (1989). 12. Molyneaux, K.A., et al. The chemokine SDF1/CXCL12 and its receptor CXCR4 regulate mouse germ cell migration and survival. Development 130, 4279-4286 (2003). 13. Hustin, J., Collette, J. & Franchimont, P. Immunohistochemical demonstration of placental alkaline phosphatase in various states of testicular development and in germ cell tumours. Int J Androl 10, 29-35 (1987). 14. Ginsburg, M., Snow, M.H. & McLaren, A. Primordial germ cells in the mouse embryo during gastrulation. Development 110, 521-528 (1990). 15. MacGregor, G.R., Zambrowicz, B.P. & Soriano, P. Tissue non-specific alkaline phosphatase is expressed in both embryonic and extraembryonic lineages during mouse embryogenesis but is not required for migration of primordial germ cells. Development 121, 1487-1496 (1995). 16. Scholer, H.R., Dressler, G.R., Balling, R., Rohdewohld, H. & Gruss, P. Oct-4: a germline-specific transcription factor mapping to the mouse t-complex. Embo J 9, 2185-2195 (1990). 21 CHAPTER 1

17. Yeom, Y.I., et al. Germline regulatory element of Oct-4 specific for the totipotent cycle of embryonal cells. Development 122, 881-894 (1996). 18. Carlson, B.M. Human embryology and developmental biology. 408-409 (2004). 19. Nef, S., et al. Testis determination requires insulin receptor family function in mice. Nature 426, 291-295 (2003). 20. Koopman, P. Sry and Sox9: mammalian testis-determining genes. Cell Mol Life Sci 55, 839- 856 (1999). 21. Capel, B., Albrecht, K.H., Washburn, L.L. & Eicher, E.M. Migration of mesonephric cells into the mammalian gonad depends on Sry. Mech Dev 84, 127-131 (1999). 22. Martineau, J., Nordqvist, K., Tilmann, C., Lovell-Badge, R. & Capel, B. Male-specific cell migration into the developing gonad. Curr Biol 7, 958-968 (1997). 23. Schmahl, J., Eicher, E.M., Washburn, L.L. & Capel, B. Sry induces cell proliferation in the mouse gonad. Development 127, 65-73 (2000). 24. Bendsen, E., et al. Number of germ cells and somatic cells in human fetal testes during the first weeks after sex differentiation. Hum Reprod 18, 13-18 (2003). 25. Bendsen, E., Byskov, A.G., Andersen, C.Y. & Westergaard, L.G. Number of germ cells and somatic cells in human fetal ovaries during the first weeks after sex differentiation. Hum Reprod 21, 30-35 (2006). 26. Sorensen, K.P., Lutterodt, M.C., Mamsen, L.S., Byskov, A.G. & Larsen, J.K. Proliferation of germ cells and somatic cells in first trimester human embryonic gonads as indicated by S and S+G(2) +M phase fractions. Cell Prolif 44, 224-233 (2011). 27. Castrillon, D.H., Quade, B.J., Wang, T.Y., Quigley, C. & Crum, C.P. The human VASA gene is specifically expressed in the germ cell lineage. Proc Natl Acad Sci U S A 97, 9585-9590 (2000). 28. Horie, K., et al. The expression of c-kit protein in human adult and fetal tissues. Hum Reprod 8, 1955-1962 (1993). 29. Jorgensen, N., Giwercman, A., Muller, J. & Skakkebæk, N.E. Immunohistochemical markers of carcinoma in situ of the testis also expressed in normal infantile germ cells. Histopathology 22, 373-378 (1993). 30. Pinkerton, J.H., Mc, K.D., Adams, E.C. & Hertig, A.T. Development of the human ovary--a study using histochemical technics. Obstet Gynecol 18, 152-181 (1961). 31. Schnieders, F., et al. Testis-specific protein, Y-encoded (TSPY) expression in testicular tissues. Hum Mol Genet 5, 1801-1807 (1996). 32. Besmer, P., et al. The kit-ligand (steel factor) and its receptor c-kit/W: pleiotropic roles in gametogenesis and melanogenesis. Dev Suppl, 125-137 (1993). 33. Strohmeyer, T., et al. Expression of the c-kit proto-oncogene and its ligand stem cell factor (SCF) in normal and malignant human testicular tissue. J Urol 153, 511-515 (1995). 34. Mauduit, C., Hamamah, S. & Benahmed, M. Stem cell factor/c-kit system in spermatogenesis. Hum Reprod Update 5, 535-545 (1999). 35. Woodward, P.J., et al. World Health Organization Classification of Tumours: Tumours of the Testis and Paratesticular Tissue. in Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs (ed. Eble J.N., S.G., Epstein J.I., Sesterhenn I.A.) 217-249 (IARC Press, Lyon, 2004). 36. Oosterhuis, J.W. & Looijenga, L.H. Testicular germ-cell tumours in a broader perspective. Nat Rev Cancer 5, 210-222 (2005). 37. Eble JN, S.G., Epstein JI, Sesterhehnn IA. Pathology and genetics of tumours of the urinary system and male genital organs. WHO classification of tumors, 217-278 (2004). 38. Post, P.N., Casparie, M.K., ten Kate, F.J. & Oosterhuis, J.W. [The epidemiology of tumors of the testes in the Netherlands: accurate rendering by the Registry of Histopathology and Cytopathology (PALGA)]. Ned Tijdschr Geneeskd 148, 1150-1154 (2004).

22 General Introduction

39. Cummings, O.W., Ulbright, T.M., Eble, J.N. & Roth, L.M. Spermatocytic seminoma: an immunohistochemical study. Hum Pathol 25, 54-59 (1994). 40. Steiner, H., et al. Metastatic spermatocytic seminoma--an extremely rare disease. Eur Urol 49, 183-186 (2006). 41. Lombardi, M., Valli, M., Brisigotti, M. & Rosai, J. Spermatocytic seminoma: review of the literature and description of a new case of the anaplastic variant. Int J Surg Pathol 19, 5-10 (2011). 42. Stoop, H., et al. Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological implications. Lab Invest 81, 919-928 (2001). 43. Rajpert-De Meyts, E., et al. The immunohistochemical expression pattern of Chk2, p53, p19INK4d, MAGE-A4 and other selected antigens provides new evidence for the premeiotic origin of spermatocytic seminoma. Histopathology 42, 217-226 (2003). 44. Goriely, A., et al. Activating mutations in FGFR3 and HRAS reveal a shared genetic origin for congenital disorders and testicular tumors. Nat Genet 41, 1247-1252 (2009). 45. Looijenga, L.H., et al. Genomic and expression profiling of human spermatocytic seminomas: primary spermatocyte as tumorigenic precursor and DMRT1 as candidate chromosome 9 gene. Cancer Res 66, 290-302 (2006). 46. Huyghe, E., Matsuda, T. & Thonneau, P. Increasing incidence of testicular cancer worldwide: a review. J Urol 170, 5-11 (2003). 47. Garner, M.J., Turner, M.C., Ghadirian, P. & Krewski, D. Epidemiology of testicular cancer: An overview. Int J Cancer (2005). 48. Huyghe, E., Plante, P. & Thonneau, P.F. Testicular cancer variations in time and space in Europe. Eur Urol 51, 621-628 (2007). 49. Schmoll, H.J., et al. Testicular non-seminoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 21 Suppl 5, v147-154 (2010). 50. Schmoll, H.J., et al. Testicular seminoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 21 Suppl 5, v140-146 (2010). 51. Horwich, A., Shipley, J. & Huddart, R. Testicular germ-cell cancer. Lancet 367, 754-765 (2006). 52. Verhoeven, R.H., Coebergh, J.W., Kiemeney, L.A., Koldewijn, E.L. & Houterman, S. Testicular cancer: trends in mortality are well explained by changes in treatment and survival in the southern Netherlands since 1970. Eur J Cancer 43, 2553-2558 (2007). 53. Fung, C. & Vaughn, D.J. Complications associated with chemotherapy in testicular cancer management. Nat Rev Urol 8, 213-222 (2011). 54. Parkin, D.M., Bray, F., Ferlay, J. & Pisani, P. Estimating the world cancer burden: Globocan 2000. Int J Cancer 94, 153-156 (2001). 55. Myrup, C., et al. Testicular cancer risk in first- and second-generation immigrants to Denmark. J Natl Cancer Inst 100, 41-47 (2008). 56. Moller, H. Decreased testicular cancer risk in men born in wartime. J Natl Cancer Inst 81, 1668-1669 (1989). 57. Hemminki, K., Li, X. & Czene, K. Cancer risks in first-generation immigrants to Sweden. Int J Cancer 99, 218-228 (2002). 58. Ekbom, A., Richiardi, L., Akre, O., Montgomery, S.M. & Sparen, P. Age at immigration and duration of stay in relation to risk for testicular cancer among Finnish immigrants in Sweden. J Natl Cancer Inst 95, 1238-1240 (2003). 59. Dieckmann, K.P. & Pichlmeier, U. Clinical epidemiology of testicular germ cell tumors. World J Urol 22, 2-14 (2004). 60. Kristensen, D.M., et al. Origin of pluripotent germ cell tumours: the role of microenvironment during embryonic development. Mol Cell Endocrinol 288, 111-118 (2008).

23 CHAPTER 1

61. Muller, J., Skakkebæk, N.E., Nielsen, O.H. & Graem, N. Cryptorchidism and testis cancer. Atypical infantile germ cells followed by carcinoma in situ and invasive carcinoma in adulthood. Cancer 54, 629-634 (1984). 62. Giwercman, A., Grindsted, J., Hansen, B., Jensen, O.M. & Skakkebæk, N.E. Testicular cancer risk in boys with maldescended testis: a cohort study. J Urol 138, 1214-1216 (1987). 63. Swerdlow, A.J., Higgins, C.D. & Pike, M.C. Risk of testicular cancer in cohort of boys with cryptorchidism. Bmj 314, 1507-1511 (1997). 64. Moller, H., Cortes, D., Engholm, G. & Thorup, J. Risk of testicular cancer with cryptorchidism and with testicular biopsy: cohort study. Bmj 317, 729 (1998). 65. Thonneau, P.F., Gandia, P. & Mieusset, R. Cryptorchidism: incidence, risk factors, and potential role of environment; an update. J Androl 24, 155-162 (2003). 66. Virtanen, H.E. & Toppari, J. Epidemiology and pathogenesis of cryptorchidism. Hum Reprod Update 14, 49-58 (2008). 67. Winter, C. & Albers, P. Testicular germ cell tumors: pathogenesis, diagnosis and treatment. Nat Rev Endocrinol 7, 43-53 (2011). 68. Hotaling, J.M. & Walsh, T.J. Male infertility: a risk factor for testicular cancer. Nat Rev Urol 6, 550-556 (2009). 69. Giwercman, A., Carlsen, E., Keiding, N. & Skakkebæk, N.E. Evidence for increasing incidence of abnormalities of the human testis: a review. Environ Health Perspect 101 Suppl 2, 65-71 (1993). 70. Moller, H., Prener, A. & Skakkebæk, N.E. Testicular cancer, cryptorchidism, inguinal hernia, testicular atrophy, and genital malformations: case-control studies in Denmark. Cancer Causes Control 7, 264-274 (1996). 71. Weidner, I.S., Moller, H., Jensen, T.K. & Skakkebæk, N.E. Risk factors for cryptorchidism and hypospadias. J Urol 161, 1606-1609 (1999). 72. Hoei-Hansen, C.E., Rajpert-De Meyts, E., Daugaard, G. & Skakkebæk, N.E. Carcinoma in situ testis, the progenitor of testicular germ cell tumours: a clinical review. Ann Oncol 16, 863- 868 (2005). 73. Che, M., et al. Bilateral testicular germ cell tumors: twenty-year experience at M. D. Anderson Cancer Center. Cancer 95, 1228-1233 (2002). 74. Coogan, C.L., et al. Bilateral testicular tumors: management and outcome in 21 patients. Cancer 83, 547-552 (1998). 75. Bokemeyer, C., et al. Bilateral testicular tumours: prevalence and clinical implications. Eur J Cancer 29A, 874-876 (1993). 76. Wilkinson, M. & Carroll, P.R. Testicular carcinoma in patients positive and at risk for human immunodeficiency virus. J Urol 144, 1157-1159 (1990). 77. Birkeland, S.A., et al. Cancer risk after renal transplantation in the Nordic countries, 1964- 1986. Int J Cancer 60, 183-189 (1995). 78. Goedert, J.J., Purdue, M.P., McNeel, T.S., McGlynn, K.A. & Engels, E.A. Risk of germ cell tumors among men with HIV/acquired immunodeficiency syndrome. Cancer Epidemiol Biomarkers Prev 16, 1266-1269 (2007). 79. Greene, M.H., et al. Familial testicular germ cell tumors in adults: 2010 summary of genetic risk factors and clinical phenotype. Endocr Relat Cancer 17, R109-121 (2010). 80. Kratz, C.P., Mai, P.L. & Greene, M.H. Familial testicular germ cell tumours. Best Pract Res Clin Endocrinol Metab 24, 503-513 (2010). 81. Rapley, E.A., et al. A genome-wide association study of testicular germ cell tumor. Nat Genet 41, 807-810 (2009). 82. Kanetsky, P.A., et al. Common variation in KITLG and at 5q31.3 predisposes to testicular germ cell cancer. Nat Genet 41, 811-815 (2009). 83. Turnbull, C., et al. Variants near DMRT1, TERT and ATF7IP are associated with testicular germ cell cancer. Nat Genet 42, 604-607 (2010).

24 General Introduction

84. Rapley, E.A., et al. Localization to Xq27 of a susceptibility gene for testicular germ-cell tumours. Nat Genet 24, 197-200 (2000). 85. Rapley, E.A., et al. Somatic mutations of KIT in familial testicular germ cell tumours. Br J Cancer 90, 2397-2401 (2004). 86. Rapley, E. Susceptibility alleles for testicular germ cell tumour: a review. Int J Androl 30, 242-250; discussion 250 (2007). 87. Rapley, E.A. & Nathanson, K.L. Predisposition alleles for Testicular Germ Cell Tumour. Curr Opin Genet Dev 20, 225-230 (2010). 88. Skakkebæk, N.E., Rajpert-De Meyts, E. & Main, K.M. Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects. Hum Reprod 16, 972-978 (2001). 89. Bosland, M.C. Hormonal factors in carcinogenesis of the prostate and testis in humans and in animal models. Prog Clin Biol Res 394, 309-352 (1996). 90. Sharpe, R.M. The 'oestrogen hypothesis'- where do we stand now? Int J Androl 26, 2-15 (2003). 91. Wohlfahrt-Veje, C., Main, K.M. & Skakkebæk, N.E. Testicular dysgenesis syndrome: foetal origin of adult reproductive problems. Clin Endocrinol (Oxf) 71, 459-465 (2009). 92. Skakkebæk, N.E. Endocrine disrupters and testicular dysgenesis syndrome. Horm Res 57 Suppl 2, 43 (2002). 93. Virtanen, H.E., Rajpert-De Meyts, E., Main, K.M., Skakkebæk, N.E. & Toppari, J. Testicular dysgenesis syndrome and the development and occurrence of male reproductive disorders. Toxicol Appl Pharmacol (2005). 94. Fisher, J.S. Environmental anti-androgens and male reproductive health: focus on phthalates and testicular dysgenesis syndrome. Reproduction 127, 305-315 (2004). 95. Heudorf, U., Mersch-Sundermann, V. & Angerer, J. Phthalates: toxicology and exposure. Int J Hyg Environ Health 210, 623-634 (2007). 96. Kang, S.C. & Lee, B.M. DNA methylation of estrogen receptor alpha gene by phthalates. J Toxicol Environ Health A 68, 1995-2003 (2005). 97. Wanderas, E.H., Grotmol, T., Fossa, S.D. & Tretli, S. Maternal health and pre- and perinatal characteristics in the etiology of testicular cancer: a prospective population- and register- based study on Norwegian males born between 1967 and 1995. Cancer Causes Control 9, 475-486 (1998). 98. Holl, K., et al. Endogenous steroid hormone levels in early pregnancy and risk of testicular cancer in the offspring: a nested case-referent study. Int J Cancer 124, 2923-2928 (2009). 99. Hughes, I.A. The quiet revolution: Disorders of sex development. Best Pract Res Clin Endocrinol Metab 24, 159-162 (2010). 100. Looijenga, L.H., et al. Gonadal tumours and DSD. Best Pract Res Clin Endocrinol Metab 24, 291-310 (2010). 101. Page, D.C. Hypothesis: a Y-chromosomal gene causes gonadoblastoma in dysgenetic gonads. Development 101 Suppl, 151-155 (1987). 102. Cools, M., Drop, S.L., Wolffenbuttel, K.P., Oosterhuis, J.W. & Looijenga, L.H. Germ cell tumors in the intersex gonad: old paths, new directions, moving frontiers. Endocr Rev 27, 468-484 (2006). 103. Looijenga, L.H., et al. Tumor risk in disorders of sex development (DSD). Best Pract Res Clin Endocrinol Metab 21, 480-495 (2007). 104. Hersmus, R., et al. FOXL2 and SOX9 as parameters of female and male gonadal differentiation in patients with various forms of disorders of sex development (DSD). J Pathol 215, 31-38 (2008). 105. Cools, M., Looijenga, L.H., Wolffenbuttel, K.P. & Drop, S.L. Disorders of sex development: update on the genetic background, terminology and risk for the development of germ cell tumors. World J Pediatr 5, 93-102 (2009).

25 CHAPTER 1

106. Wilms, M. Die tertoiden Geschwulste des Hodens, mit Einschluss der sog. Cystoide und Enchondrome. Beitrage der Pathologische Anatomie 19, 233-366 (1896). 107. Azzopardi, J.G., Mostofi, F.K. & Theiss, E.A. Lesions of testes observed in certain patients with widespread choriocarcinoma and related tumors. The significance and genesis of hematoxylin-staining bodies in the human testis. Am J Pathol 38, 207-225 (1961). 108. Skakkebæk, N.E. Possible carcinoma-in-situ of the testis. Lancet 2, 516-517 (1972). 109. Ewing, J. Teratoma testis and its derivatives. Surgical Gynaecology and Obstetrics 12, 230- 261 (1911). 110. Scully, R. Intratubular germ cell neoplasia (carcinoma in situ). World Urology Update Series 1, 1-8 (1982). 111. Rosai, J. Ackerman's surgical pathology 8th edition Mosby; St Louise, 1279 (1996). 112. Loy, V. & Dieckmann, K.P. Carcinoma in situ of the testis: intratubular germ cell neoplasia or testicular intraepithelial neoplasia? Hum Pathol 21, 457-458 (1990). 113. Almstrup, K., et al. Embryonic stem cell-like features of testicular carcinoma in situ revealed by genome-wide gene expression profiling. Cancer Res 64, 4736-4743 (2004). 114. Skotheim, R.I., et al. Differentiation of human embryonal carcinomas in vitro and in vivo reveals expression profiles relevant to normal development. Cancer Res 65, 5588-5598 (2005). 115. Novotny, G.W., et al. Analysis of gene expression in normal and neoplastic human testis: new roles of RNA. Int J Androl 30, 316-326; discussion 326-317 (2007). 116. van de Geijn, G.J., Hersmus, R. & Looijenga, L.H. Recent developments in testicular germ cell tumor research. Birth Defects Res C Embryo Today 87, 96-113 (2009). 117. Rajpert-de Meyts, E. & Hoei-Hansen, C.E. From gonocytes to testicular cancer: the role of impaired gonadal development. Ann N Y Acad Sci 1120, 168-180 (2007). 118. Looijenga, L.H., Gillis, A.J., Stoop, H., Biermann, K. & Oosterhuis, J.W. Dissecting the molecular pathways of (testicular) germ cell tumour pathogenesis; from initiation to treatment-resistance. Int J Androl (2011). 119. von der Maase, H., et al. Carcinoma in situ of contralateral testis in patients with testicular germ cell cancer: study of 27 cases in 500 patients. Br Med J (Clin Res Ed) 293, 1398-1401 (1986). 120. Jorgensen, N., Muller, J., Giwercman, A. & Skakkebæk, N.E. Clinical and biological significance of carcinoma in situ of the testis. Cancer Surv 9, 287-302 (1990). 121. Scully, R.E. Gonadoblastoma; a gonadal tumor related to the dysgerminoma (seminoma) and capable of sex-hormone production. Cancer 6, 455-463 (1953).

A:IM>KÙ

<@F4A7BHG?

TGCT is by far the most common malignancy in Caucasian men aged 15-45 years, and the second cause of death in this age group. The combination of the rising incidence, identification of risk populations, knowledge of the precursor lesions, as well as the excellent prognosis of the patients, especially when diagnosed at an early stage of development, makes this an interesting type of cancer for early diagnosis and subsequent treatment. This will benefit the quality of life of the individual patient because it will prevent known long term sequalae of systematic treatment of the patients with invasive (metastatic) cancer using either chemotherapy or irradiation.

Within the Netherlands currently about 750 new TGCT patients are diagnosed per year, which has doubled since 1989. Histologically and clinically, TGCTs are subdivided into seminomatous and nonseminomatous tumors. The nonseminomas are further subclassified into embryonal carcinoma (being the stem cell population), teratoma (representing somatic lineages), yolk sac tumor and choriocarcinoma (representing extra-embryonic lineages). These can be pure or intermixed, even together with a seminoma component. Both seminomas and nonseminomas originate from a common precursor, known as CIS (IGCNU or TIN), which has characteristics of a primordial germ cell (PGC)/gonocyte. Indeed, a disturbed early testicular development (referred to as reduced testicularization) is a risk factor for development of TGCTs. Therefore understanding normal testicular development is of importance to shed light on the earliest events in the pathogenesis of TGCTs. For this reason, a series of normal male embryonic, fetal, and post-natal gonads are investigated, especially related to the expression of a number of proteins informative to distinguish the different maturation stages of germ cells. The results are described in Chapter 3, in which it is shown that CIS very likely finds its origin in germ cells which are arrested in development during the early embryogenesis, at the stage of PGC/gonocytes. The findings determine the window of opportunity for early diagnosis.

The possible histological composition of TGCTs indicates a pluripotent nature of the cell of origin. To further substantiate this hypothesis, it is investigated whether the tumors can also generate the germ cell lineage itself. Based on the application of a set of markers that have become available recently, supportive evidence is presented in Chapter 4. TGCTs are indeed omnipotent/totipotent, and are able to form all differentiation lineages as found during embryogenesis, including the germ line.

The pluripotent (omnipotent/totipotent) nature of TGCTs initiated a series of studies in which markers known to be relevant to regulate maintenance of pluripotency in embryonic stem cells are investigated for their staining pattern in TGCTs. This especially for the earliest identified factor OCT3/4, also known as

28 AIMS AND OUTLINES

POU5F1. To study the protein expression profile in normal gonadal development and tumors, immunohistochemistry is applied, both on individual sections, as well as tissue micro arrays. The results, as described in Chapter 5 and 6, demonstrate that OCT3/4 is an absolute marker for detection of the precursor lesions CIS and GB, seminomatous tumors and embryonal carcinoma. All tumor cells show a strong positive staining within the nucleus, while in the embryonal carcinoma cells also a cytoplasmic localization is identified. Upon differentiation, OCT3/4 is lost. Based on this highly reproducible finding, both using monoclonal and polyclonal antibodies, it is investigated whether OCT3/4 can also be used as marker for development of a non-invasive test, i.e., detection of CIS cells in semen. In Chapter 7, the results of the first study are presented showing that in principle CIS cells can be identified in semen using immunohistochemical detection of OCT3/4. The diagnostic application of OCT3/4 is not limited to TGCTs. In Chapter 8 it is shown that OCT3/4 can also be used to diagnose extragonadal seminomatous tumors, including those located in the brain.

In Chapter 9 the transcription factors SOX2 and SOX17 are shown to be very useful in the diagnostics of TGCTs. During embryogenesis they interact with OCT3/4, important for normal embryogenesis. Including these markers in the diagnostic workup of TGCTs, next to OCT3/4, will lead to an optimal diagnosis, and subsequent treatment of the patient. Patients with cryptorchidism and more severe DSD are high risk groups for the development of a TGCT. Diagnosis at young age in these cases is of eminent importance. A gonadal biopsy can give important information, however, especially in the forementioned cases, germ cell maturation delay can be present. This will result in an extended time window of expression of embryonic markers which are used for CIS detection in adults. In Chapter 10 the value of the immunohistochemical detection of KITLG (SCF) for this group of young individuals is demonstrated. Early detection will benefit quality of life. In Chapter 11 an enzymehistochemical method is described which makes it possible to examine frozen sections for the presence of CIS, seminoma or embryonal carcinoma. This method makes it possible to guide the decision, during the initial surgical intervention, to remove the gonad, or perform limited surgery.

During development of the gonads, especially related to the germ cell population, changes in DNA methylation by means of cytosine methylation, is a crucial regulatory mechanism for gene expression. Most investigations, also on TGCTs, are done using isolated DNA, which do not allow analysis of the cells under investigation in their natural environment. This is solved by means of immunohistochemistry, of which results are described in Chapter 12. PGCs/gonocytes are showing an overall demethylated pattern, while spermatogonia and later germ cell maturation stages become more methylated. 29 CHAPTER 2

CIS, GB, and seminomas are also hypomethylated, while the differentiated nonseminomatous elements are hypermethylated and embryonal carcinoma show an intermediate pattern. Cisplatin refractory seminoma shows a hypermethylated pattern. Cell line investigations support the model that cFLAR (also known as cFLIP) might be a regulator of cisplatin sensitivity in TGCTs. In Chapter 13, the results obtained are discussed in a broader perspective regarding their possible impact in the diagnostic workup of TGCTs.

A:IM>KÚ

#4G;B58EFB9G8FG<6H?4E 64E6

BHEA4?B9#4G;B?B:LÒÚÓ ÙÔØ Ùß× ÒÚÚÔ

E<878@4AABA86>8E 4AF&GBBC %BA4?7%78E<=:8E ,HA 4<;E8@8L8E 88A78EG BB<=8A:4 ';8F84HG;BEF6BAGE<5HG87 8DH4??LGBG; CHAPTER 3

Abstract Several proteins, such as the placental/germ cell alkaline phosphatases (PLAPs), the stem cell factor receptor c-KIT, and the transcriptional regulator and marker of pluripotency OCT3/4, have been found in both normal immature and malignant germ cells, known as carcinoma in situ/intratubular germ cell neoplasia unclassified (CIS/ITGCNU). In the present study, immunohistochemical methods were used to evaluate the expression of these markers in a series of male gonads from fetuses from the second and third trimesters, and neonates. In addition to these markers, the presence of VASA (a protein specific for the germ cell lineage), TSPY (the testis-specific protein, Y-encoded), and the proliferation index (Ki-67 antigen) was analyzed. All these proteins are reported to be present both during spermatogenesis and in CIS/ITGCNU. Positive staining for VASA with varying intensity was found in all germ cells, while TSPY was predominantly located in the pre-spermatogonial cells at all developmental ages. In contrast, the markers PLAP, c-KIT, OCT3/4, and Ki-67 were more frequent at earlier developmental stages and decreased gradually with time, although they could occasionally be detected in germ cells of neonates. These findings are in line with a declining number of gonocytes during fetal development, concomitant with an increase in the number of pre-spermatogonia. The latter have lost the immature germ cell phenotype. These findings are compatible with the hypothesis that CIS/ITGCNU arises from developmentally arrested germ cells, most likely primordial germ cells/gonocytes, at an early time point during intrauterine development.

32 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS Introduction In humans and rodents, amongst other species, the embryonic precursors of the gametes are known as primordial germ cells (PGCs) (see ref 1 for a review). These cells are set aside to an extra-embryonic localization early during embryonic development and around the fifth and sixth weeks of human development, they migrate to the area where the genital ridge will be formed [2]. Subsequently, gonadal and sexual differentiation occur in the sixth and seventh weeks [3]. With regard to testicular development, the germ cells at this particular stage are referred to as gonocytes, which are predominantly found in the central areas of the newly formed tubules. Starting around the 14th week of development, these gonocytes gradually migrate towards the tubular periphery. Once the cells are in close contact with the basal lamina of the tubule, they are referred to as pre- spermatogonia. Maturation is a gradual process; at the 20th gestational week, the fetal testis predominantly contains pre-spermatogonial germ cells [4]. During the first few years after birth and until puberty, morphological and functional changes occur, including a change from large, immature germ cells (pre- spermatogonia) to adult type A spermatogonia [5–7]. With the onset of puberty, spermatogonia undergo further spermatogenic maturation and after meiotic divisions, finally produce spermatozoa.

On the basis of multiple findings, it has been hypothesized that CIS/ITGCNU, the common precursor lesion of adult testicular germ cell tumours (TGCTs), originates early during fetal development (see ref 8 for a review). This is illustrated by the presence of a number of markers common to CIS/ITGCNU and immature germ cells, including germ cell/placental alkaline phosphatases (PLAPs), the proto-oncogene receptor c-KIT, and the transcriptional regulator and marker of pluripotency OCT3/4. CIS/ITGCNU cells phenotypically and ultrastructurally resemble PGCs/gonocytes [9–11]. Moreover, epidemiological data support the hypothesis that the initiating event of TGCT development occurs during the fetal period [12]. In spite of these data, an alternative origin of TGCTs has been suggested, in which the pachytene spermatocyte, a cell not found during the prepubertal period, has been suggested to be the target of transformation [13].

In order to investigate the origin of CIS/ITGCNU further, we undertook an extensive study of the presence of a number of markers during normal fetal testicular development. This will help us to define further the emerging phenotype of CIS/ITGCNU and shed light on the possible limitations of these markers for early diagnosis of CIS/ITGCNU in high-risk neonates and infants.

33 CHAPTER 3 Material and methods

Tissue samples Use of tissues for scientific reasons was approved by an institutional review board (MEC 02.981). The samples were used according to the ‘Code for Proper Secondary Use of Human Tissue in The Netherlands’, as developed by the Dutch Federation of Medical Scientific Societies (FMWV) (version 2002).

Human fetal gonads from 27 males from the second and third trimesters after spontaneous or induced abortions (gestational age 15–40 weeks), or from premature and term neonates that had died shortly after birth, and one infant who died from cot death at the age of 6 weeks were obtained from post-mortem sections in our department. Testes were dissected and fixed in 10% formalin and processed into paraffin wax. To ensure satisfactory quality, poor preservation of tissue samples assessed by haematoxylin and eosin staining led to exclusion from this analysis. Cases showing conditions that possibly affect gonadal development, such as trisomy 13, 18 and 21, hydrocephalus, maldeveloped kidneys, or gross intrauterine growth retardation, were excluded from the study. Gestational ages were calculated in relation to the mother’s last menstrual cycle and were in accordance with the foot length and the crown–heel length measurements at autopsy, showing a maximum variation of 2 weeks.

Table 1. Antibodies (source) and detection method used for immunohistochemistry

Secondary antibody Pre- (1 : 200)(biotinylated) Antibody Company Code treatment Dilution Code Visualization

* Heat-induced antigen retrieval [42]. † ABC complex, alkaline phosphatase, Dako Code: K0391. ‡ ABC complex, horseradish peroxidase, Dako Code: K0377.

34 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS Histochemical and immunohistochemical staining For immunohistochemistry, sections were incubated with the primary antibody overnight at 40C (PLAP, c-KIT, TSPY, VASA) or for 2 h at room temperature (OCT3/4, Ki-67). The antibodies used are indicated in Table 1. Sections were counterstained with haematoxylin.

Double staining was performed by using a combination of the same detection method but with different substrates: Fast blue/naphthol ASMX phosphate (F3378 and N500; Sigma, Steinheim, Germany) for blue staining and 3-amino-9- ethyl-carbazole (A.5754 and D4254; Sigma, Steinheim, Germany)/H2O2 for red staining, without counterstain. Endogenous peroxidase activity and/or endogenous biotin was blocked using 3% H2O2 (for 5 min) and/or a blocking kit for endogenous biotin (Vector Laboratories, Burlingame, CA, USA) to prevent background staining [14].

For quantification, cell numbers showing a positive signal were counted in five to ten cross-sections of seminiferous tubules by two different observers (FH and HS) who were blinded to the gestational age at which the material was sampled.

Results The presence of a number of well-known and novel markers for CIS/ITGCNU during normal fetal testicular development was studied. The available data on these markers are summarized in Table 2. The results are reported separately for the known CIS/ITGCNU markers (PLAP, c-KIT, and OCT3/4) and the other markers analysed (VASA, TSPY, and Ki-67). Subsequently, the results of double staining are described. Representative illustrations are shown in Figure 1. Figure 2 summarizes the results of each marker for each case individually. For the majority of cases, intra-individual comparison of the expression of CIS/ITGCNU markers (PLAP, c-KIT, and OCT3/4) gave consistent results, reflecting the state of maturation of one particular case. The data for the whole population are depicted graphically in Figure 3 (grey bars) and compared with findings published so far (black bars).

35 CHAPTER 3

Table 2. Overview of markers for carcinoma in situ/intratubular germ cell neoplasia unclassified (CIS/ITGCNU), proliferation, and germ cell-specific factors

Marker/ CIS/ITGCNU Antigen (intensity) References

Glycogen +++ Nielsen and Lein, 1974 [20] PLAP +++ Beckstead, 1983 [44]; Jacobsen and Norgaard-Pedersen, 1984 [45] c-KIT +++ Rajperts-De Meyts,d Skakkebæk, 1994 [46];Strohmeyer et al, 1995 [47] OCT3/4 +++ Looijenga et al, 2003 [27] VASA + Zeeman et al, 2002 [48] TSPY ++ Schnieders et al, 1996 [38]; Lau et al, 2000 [39] Ki-67 +++ Datta et al, 2000 [30]

Immunohistochemical detection of PLAP, c-KIT, and OCT3/4 Positive staining for PLAP, c-KIT, and OCT3/4 was seen specifically in the germ cells, while no staining was present in Sertoli, Leydig, or interstitial cells (see Figure 1).

The highest number of germ cells staining for PLAP was seen in the earliest stages of fetal development examined and decreased continuously through-out the following weeks with advancing gestational age. Within the tubules of earlier stages, PLAP was predominantly detected in gonocytes, and to a lesser extent in pre-spermatogonia (see Figure 1A). After birth, PLAP could still be found occasionally (see Figure 2A), with maximally one positive cell per visual field. These cells were almost exclusively located in the centre of the tubules. Mainly during the second trimester, a large number of germ cells were positive for c-KIT, which, like PLAP, declined gradually throughout gestation. Again, positive germ cells were detectable at term, albeit at low numbers. Within the tubules, c-KIT was seen in both gonocytes and pre-spermatogonia (see Figures 1B and 2B).

OCT3/4 resulted in nuclear staining of germ cells at all gestational ages. With advancing gestational age, a constant decrease in cells staining for OCT3/4 was found (see Figure 2C). High numbers with an average of four to six cells per tubule, mainly gonocytes, were seen throughout the first half of the second trimester (Figure 1C). Throughout the second half of the second trimester, the average number of positive cells decreased to less than three cells per tubule. At term, only a few positive germ cells were detectable, mainly located in the centre of the tubules.

36 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS An increased number of germ cells positive for PLAP, c-KIT, and OCT3/4 were seen in testes from three second-trimester fetuses with chromosomal abnormalities, ie trisomy 21 and 18: these were not included in the series presented (data not shown). These findings are in line with previous reports [15,16] and support the model that chromosomal abnormalities can interfere with normal germ cell maturation.

Immunohistochemical detection of VASA, TSPY, and Ki-67 VASA-positive germ cells were found at all gestational ages and after birth. Although the staining intensity was variable, pre-spermatogonia showed consistently stronger staining than gonocytes (see Figure 1D). The number of positive cells per tubule decreased only slightly with gestational age, resulting in a different overall staining pattern compared with the factors described above: VASA was still found in germ cells at term and in neonates (see Figure 2D). Nuclear and cytoplasmic TSPY was predominantly observed in putative pre- spermatogonia, based on their peripheral localization within the tubules, at all gestational ages. No decrease in staining was found at term or in the first few weeks after birth (see Figure 2E). On average, three to five cells per tubule showed positive staining, most often seen in pairs or groups of germ cells (Figure 1E). In contrast to the other markers described so far, TSPY was not restricted to germ cells, but was also detected in Leydig cells at all ages examined.

Ki-67 showed nuclear staining in a high number of cells, both within and outside the tubules, throughout the whole period of testicular development investigated. In developing tubules, until 24 weeks of gestational age, Ki-67 was predominantly seen in gonocytes (Figure 1F); after 24 weeks, both basally and centrally located germ cells showed staining in roughly equal numbers. The number of positive cells decreased steadily with advancing age, and perinatally, only a few intratubular cells in the tubules remained positive (see Figure 2F). These were mostly centrally located in the tubules, whereas the majority of pre- spermatogonia seemed to have entered a quiescent phase at that time.

Results of double staining Double staining was performed to assess correlations or differences in the presence of various markers in germ cells at different gestational ages. A total of seven cases at 15, 18, 21, 24, 27, 30, and 35 weeks’ gestational age were stained for four different combinations: OCT3/4 and PLAP; OCT3/4 and c-KIT; PLAP and c- KIT; and OCT3/4 and VASA (Figures 1G–1J, respectively). The results of all double- staining experiments were in accordance with the single staining results. Staining for OCT3/4 and PLAP revealed that at all stages, a higher number of germ cells were positive for OCT3/4 than for PLAP. PLAP was never detected in germ cells negative for OCT3/4 and the presence of OCT3/4 was more frequently

37 CHAPTER 3 observed at later stages of gestational development, when PLAP was almost undetectable. In contrast, OCT3/4 and c-KIT were found in comparable numbers of germ cells at earlier stages (second trimester), but c-KIT remained positive in more germ cells than OCT3/4 at later stages (third trimester). In agreement with these findings, double staining for PLAP and c-KIT showed that at all gestational ages more germ cells were positive for c-KIT than for PLAP. PLAP was never detected in germ cells negative for c-KIT, but due to an earlier loss of PLAP, the ratio of cells showing only c-KIT expression versus the number of cells positive for both factors increased from approximately 3 : 1 in the second trimester to approximately 9 : 1 during the third trimester.

An inverse correlation was found between the presence of OCT3/4 and the intensity of VASA. Gonocytes were positive for OCT3/4, but only weakly positive for VASA during the second trimester, whereas pre-spermatogonia staining strongly for VASA mostly lacked OCT3/4 (Figure 1J).

Figure 1. Results of the immunohistochemical analysis of different markers in fetal germ cells. (A) Fetal testis (15 weeks of development); PLAP (red signal) is seen in gonocytes and pre- spermatogonia. (B) Same case; c-KIT staining (red signal) is seen in a large number of gonocytes and pre-spermatogonia. (C) Same case; OCT3/4 (brown nuclear signal). (D) Same case; VASA (red signal). Note the variation in staining, with weaker signal intensity in gonocytes (more centrally in tubules) than in pre-spermatogonia (on the basal membrane of tubules). (E) Testis of a neonate (6 weeks); TSPY (red nuclear and cytoplasmic signal), mainly in pre-spermatogonia. Note also the nuclear and cytoplasmic staining of Leydig cells (arrow). (F) Fetal testis, 24 weeks: Ki-67 (brown nuclear signal) is seen both in tubules and in interstitial cells. (G) Fetal testis, 21 weeks: double staining for OCT3/4 (red nuclear signal) and PLAP (blue cytoplasmic signal). OCT3/4 and PLAP are co-expressed in the majority of immature germ cells, and occasional cells are positive for OCT3/4 but negative for PLAP (arrowhead). Note higher magnification. (H) Fetal testis, 15 weeks: double staining for OCT3/4 (blue nuclear signal) and c-KIT (red membranous signal). Note co-expression of OCT3/4 and c-KIT in the majority of germ cells. (I) Fetal testis, 15 weeks: double staining for PLAP (blue cytoplasmic signal) and c-KIT (red membranous signal). Note that co-expression of both markers results in a dark, almost black signal. The majority of germ cells express only c-KIT, whereas only a minority are positive for both markers. (J) Fetal testis, 21 weeks: double staining for OCT3/4 (red nuclear signal) and VASA (blue cytoplasmic signal). Germ cells showing a strong intensity for VASA are negative for OCT3/4 and are more often found on the basal lamina of the tubule. Note higher magnification.

38 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS

39 CHAPTER 3 Discussion The aim of this study was to examine the differential presence of a number of factors known to be present in both human fetal germ cells and CIS/ITGCNU, the pre-invasive stage of TGCTs. In addition, we studied a number of factors associated with the proliferation and differentiation of germ cells, allowing us to correlate our findings with the fate of these cells during the second and third trimesters and the first weeks after birth.

Overall, the presence of all CIS/ITGCNU markers decreases during the second and third trimesters (see Figure 2, left panels). Nevertheless, with increasing age, expression was down-regulated differentially. This resulted in a specific expression pattern for each of these factors during fetal male germ cell development. We interpret these data in the context of a gradual process of maturation, extending into the first year of life, which is in accordance with earlier reports [17–20]. The results of the different double-staining experiments indicate that PLAP is the first factor that is down-regulated during the development of normal germ cells, followed by OCT3/4 and finally c-KIT. The fact that CIS/ITGCNU shows high expression of PLAP points towards an early origin of the precursor cell during intrauterine development. In the following paragraphs, the most interesting findings of the individual markers included in this study will be discussed in detail, predominantly in the context of understanding the pathobiological consequences of our findings.

c-KIT is a type III receptor tyrosine kinase, of which stem cell factor (SCF) is the ligand (see ref 21 for a review). The c-KIT/SCF system has been found to be involved in survival and proliferation of migrating germ cells in mice [22]. Our recent finding of the presence of activating mutations affecting one specific site of the receptor in the majority of bilateral TGCTs is in agreement with this model and supports activation of c-KIT as an early initiation event in the pathogenesis of TGCT [23]. Here we show a high presence of c-KIT at the early stages of germ cell development. Its expression generally declines with advancing age during the intrauterine period, but persists throughout the second and third trimesters and to a lesser extent after birth. This is in contrast to earlier reports, in which the authors concluded from a rapid decrease of c-KIT at 10–13 weeks that malignant transformation takes place early during fetal development, possibly even before week 10 [10]. However, more in line with the observation of this study, the detection of c-KIT in germ cells at later stages of development has been described in a recent report [24]. The discrepancies between the different studies might be due to differences in the sensitivity of the antibodies and the detection methods used.

40 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS OCT3/4 is a transcriptional regulator, exclusively found in pluripotent human and mouse embryonic stem cells and early germ cells (see refs 25–28 for a review). Recently, it has been described to be present in specific subtypes of TGCTs, including CIS/ITGCNU [27,29], in the context of which it is an informative diagnostic marker [27]. Here we show, in contrast to our earlier more limited study, that OCT3/4 can still be present during the first weeks after birth, albeit at very low frequency. In neonates, OCT3/4 is almost exclusively expressed in gonocyte-like cells located centrally in the tubules, and not in pre-spermatogonia.

Ki-67 is a nuclear antigen, which is studied in CIS/ITGCNU and TGCTs [30,31]. This antigen is informative in the assessment of proliferation, including germ cells in normal human testis, both during the fetal period and in newborns. Conflicting data on Ki-67 in germ cells at term have been reported [7,32]. To define further the population of proliferating germ cells during the intrauterine period, we assessed the expression of Ki-67 at different gestational ages. Throughout the second trimester, there was a decrease in the number of positive cells, which remained constant at a somewhat lower level at later ages. Interestingly, proliferation was seen in two different populations of germ cells at different time points. Ki-67 was predominantly seen in gonocytes until the end of the second trimester (ie around weeks 24–26). Thereafter, both basally and centrally located cells expressed Ki-67 in roughly equal numbers, suggesting that, in addition to gonocytes, the pool of more differentiated pre-spermatogonia also expands during the late second and third trimesters, extending into the first weeks after birth.

Human TSPY, the testis-specific protein, Y-encoded, gene, is organized as a repetitive gene family mapped to the critical region of the gonadoblastoma (GBY) [33–35] locus on the short arm of the Y chromosome [36,37]. It is mainly found during early spermatogenesis and has been suggested to play a normal role in spermatogonial proliferation and an oncogenic role in early germ cell tumourigenesis [38,39]. Expression of TSPY sequences has been found by RNA analysis in prenatal and adult testes [37,40,41]. To our knowledge, we describe for the first time the presence of TSPY protein in human germ cells during male gonadal development. Staining was often seen in groups of pre-spermatogonia throughout the second and third trimesters and in neonates. Although little is known about the function of this protein, a role in the regulation of the proliferation of germ cells is tempting, as has been suggested before [39].

41 CHAPTER 3

Figure 2. Absolute numbers of cells per tubule showing positive immunohistochemical staining for the markers at different gestational ages (weeks). The earliest case studied in this series was 15 weeks; the oldest was a neonate of 6 weeks. The arrow indicates the time of birth (median 40 ± 2 weeks). The left panels (A–C) show markers for CIS/ITGCNU; the right panels (D–F) illustrate factors associated with germ cell differentiation and proliferation. Each black spot represents one case

42 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS

Figure 3. Overview of the data currently available from the literature on the expression of different antigens in germ cells during intrauterine development (black bars) compared with the findings of this study (grey bars). The sizes of the grey bars schematically represent the frequency of germ cells expressing the individual factor at different developmental ages.

In summary, our data indicate that during the second trimester, gonocytes are positive for a number of CIS/ITGCNU markers, such as OCT3/4, c-KIT, and PLAP. During the process of maturation towards pre-spermatogonia, these cellsgradually lose these early markers. During further development, there is a relative decrease in the number of gonocytes compared with more mature pre- spermatogonia. Interestingly, the time point at which more and more germ cells become attached to the basal lamina of the tubule (between weeks 20 and 24) coincides with an overall decrease in Ki-67 expression, down-regulation of the CIS/ITGCNU markers, and an increase in VASA staining intensity.

This suggests a preference for differentiation over proliferation at the transition from gonocytes to pre-spermatogonia during the second trimester and indicates that attachment to the basal membrane could be important for germ cell maturation. This model is supported by the results of the double-staining experiments for OCT3/4 and VASA, where gonocytes in the more central areas of developing tubules show expression of OCT3/4, but low staining intensity for VASA. Although dedifferentiation and consecutive re-expression of early markers cannot be ruled out as a mechanism for the development of CIS/ITGCNU, our data are in line with the model of a maturation arrest of immature germ cells as one of the first pathogenetic hits in the development of TGCTs.

43 CHAPTER 3 Two observations from the study presented are particularly noteworthy. First, the presence of markers such as PLAP, c-KIT, and OCT3/4 is not restricted to the early stages of germ cell maturation, but extends well into the second and third trimesters and can, in fact, still be found in neonates. Therefore, in contrast to the situation in the testes of adolescents and adults, these markers can be unreliable for the detection of CIS/ITGCNU in very young children. This adds further evidence to the notion that testicular biopsy is of limited value in this age group [49,50]. Second, gonocytes that are positive for these markers at the later stages of normal development are hardly ever found on the basal membrane of the seminiferous tubules, but are localized more centrally. This distinguishes them from CIS/ITGCNU cells, which are always in contact with the basal membrane and phenotypically and ultrastructurally resemble pre-spermatogonia [11]. This indicates that CIS/ITGCNU cells, possibly due to a maturation arrest, show a certain marker profile that normal germ cells in this localization have already lost. However, while retaining features of immature germ cells, CIS/ITGCNU cells also show some potential to develop along the germ cell lineage, as is documented by the presence of VASA and TSPY in these cells.

Acknowledgements We thank Diego Castrillon, Women’s and Perinatal Pathology Division, Department of Pathology (DHC), Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts, for kindly providing the polyclonal VASA antibody. This work was supported by the Dutch Cancer Society (HS, LHJL) and the Deutsche Krebshilfe, Dr Mildred Scheel Stiftung (FH).

References 1 Donovan PJ. The germ cell the mother of all stem cells. Int J Dev Biol 1998; 42: 1043–1050. 2. Witschi E. Migration of the germ cells of the human embryos from the yolk sac to the primitive gonadal folds. Contrib Embryol 1948; 209: 67–80. 3. Falin LI. The development of genital glands and the origin of germ cells in human embryogenesis. Acta Anat (Basel) 1969; 72: 195–232. 4. Gondos B, Hobel CJ. Ultrastructure of germ cell development in the human fetal testis. Z Zellforsch Mikrosk Anat 1971; 119: 1–20. 5. Muller J, Skakkebæk NE. Quantification of germ cells and seminiferous tubules by stereological examination of testicles from 50 boys who suffered from sudden death. Int J Androl 1983; 6: 143–156. 6. Chemes HE. Infancy is not a quiescent period of testicular development. Int J Androl 2001; 24: 2–7. 7. Berensztein EB, Sciara MI, Rivarola MA, Belgorosky A. Apopto-sis and proliferation of human testicular somatic and germ cells during prepuberty: high rate of testicular growth in newborns medi-ated by decreased apoptosis. J Clin Endocrinol Metab 2002; 87: 5113–5118. 8. Rajpert-De Meyts E, Jørgensen N, Brondum-Nielsen K, Muller J, Skakkebæk NE. Developmental arrest of germ cells in the pathogenesis of germ cell neoplasia. APMIS 1998; 106: 198–204; discussion 204–196.

44 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS

9. Rajpert-De Meyts E, Bartkova J, Samson M, et al. The emerging phenotype of the testicular carcinoma in situ germ cell. APMIS 2003; 111: 267–278; discussion 278–269. 10. Jørgensen N, Rajpert-De Meyts E, Graem N, Muller¨ J, Giwer-cman A, Skakkebæk NE. Expression of immunohistochemical markers for testicular carcinoma in situ by normal fetal germ cells. Lab Invest 1995; 72: 223–231. 11. Gondos B. Ultrastructure of developing and malignant germ cells. Eur Urol 1993; 23: 68–75. 12. Møller H. Decreased testicular cancer risk in men born in wartime. J Natl Cancer Inst 1989; 81: 1668–1669. 13. Chaganti RSK, Houldsworth J. The cytogenetic theory of the pathogenesis of human adult male germ cell tumours. APMIS 1998; 106: 80–84. 14. Stoop H, Van Gurp RHJLM, De Krijger R, et al. Reactivity of germ cell maturation stage- specific markers in spermatocytic seminoma: diagnostic and etiological implications. Lab Invest 2001; 81: 919–928. 15. Satge D, Sasco AJ, Cure H, Leduc B, Sommelet D, Veke-mans MJ. An excess of testicular germ cell tumors in Down’s syndrome. Three case reports and a review of the literature. Cancer 1997; 80: 929–935. 16. Hawkins E, Heifetz SA, Giller R, Cushing B. The prepubertal testis (prenatal and postnatal): its relationship to intratubular germ cell neoplasia: a combined Pediatric Oncology Group and Children’s Cancer Study Group. Hum Pathol 1997; 28: 404–410. 17. Jørgensen N, Giwercman A, Muller¨ J, Skakkebæk NE. Immuno-histochemical markers of carcinoma in situ of the testis also expressed in normal infantile germ cells. Histopathology 1993; 22: 373–378. 18. McKay DG, Hertig AT, Adams EC, Danziger S. Histochemical observations on the germ cells of human embryos. Anat Rec 1953; 117: 201–219. 19. Fujimoto T, Miyayama Y, Fuyuta M. The origin, migration and fine morphology of human primordial germ cells. Anat Rec 1977; 188: 315–330. 20. Nielsen SW, Lein DH. Tumours of the testis. Bull Wld Hlth Org 1974; 50: 71–78. 21. De Miguel MP, Cheng L, Holland EC, Federspiel MJ, Dono-van PJ. Dissection of the c-Kit signaling pathway in mouse primordial germ cells by retroviral-mediated gene transfer. Proc Natl Acad Sci U S A 2002; 99: 10 458–10 463. 22. McLaren A. Development of primordial germ cells in the mouse. Andrologia 1992; 24: 243– 247. 23. Looijenga LHJ, De Leeuw PJC, Van Oorschot M, et al. Stem cell factor receptor (c-KIT) codon 816 mutations predict development of bilateral testicular germ cell tumors. Cancer Res 2003; 63: 7674–7678. 24. Helal MA, Mehmet H, Thomas NS, et al. Ontogeny of human fetal testicular apoptosis during first, second, and third trimesters of pregnancy. J Clin Endocrinol Metab 2002; 87: 1189–1193. 25. Goto T, Adjaye J, Rodeck CH, Monk M. Identification of genes expressed in human primordial germ cells at the time of entry of the female germ line into meiosis. Mol Hum Reprod 1999; 5: 851–860. 26. Hansis C, Grifo JA, Krey LC. Oct-4 expression in inner cell mass and trophectoderm of human blastocysts. Mol Hum Reprod 2000; 6: 999–1004. 27. Looijenga LHJ, Stoop H, De Leeuw PJC, et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003; 63: 2244–2250. 28. Pesce M, Scholer HR. Oct-4: control of totipotency and germline determination. Mol Reprod Dev 2000; 55: 452–457. 29. Sperger JM, Chen X, Draper JS, et al. Gene expression patterns in human embryonic stem cells and human pluripotent germ cell tumors. Proc Natl Acad Sci U S A 2003; 100: 13 350– 13 355.

45 CHAPTER 3

30. Datta MW, Renshaw AA, Dutta A, Hoffman MA, Loughlin KR. Evaluation of cyclin expression in testicular germ cell tumors: cyclin E correlates with tumor type, advanced clinical stage, and pulmonary metastasis. Mod Pathol 2000; 13: 667–672. 31. Bartkova J, Rajpert-De Meyts E, Skakkebæk NE, Lukas J, Bartek J. Deregulation of the G1/S-phase control in human testicular germ cell tumours. APMIS 2003; 111: 252–265; discussion 265–256. 32. Ketola I, Toppari J, Vaskivuo T, Herva R, Tapanainen JS, Heikin-heimo M. Transcription factor GATA-6, cell proliferation, apoptosis, and apoptosis-related proteins Bcl-2 and Bax in human fetal testis. J Clin Endocrinol Metab 2003; 88: 1858–1865. 33. Page DC. Hypothesis: a Y-chromosomal gene causes gonadoblastoma in dysgenetic gonads. Development 1987; 101(Suppl): 151–155. 34. Tsuchiya K, Reijo R, Page DC, Discheche CM. Gonadoblastoma: molecular definition of the susceptibility region on the Y chromosome. Am J Hum Genet 1995; 57: 1400–1407. 35. Salo P, Kaariainen H, Petrovic V, Peltomaki P, Page DC, de la Chapelle A. Molecular mapping of the putative gonadoblastoma locus on the Y chromosome. Genes Chromosomes Cancer 1995; 14: 210–214. 36. Arnemann J, Epplen JT, Cooke HJ, Sauermann U, Engel W, Schmidtke J. A human Y- chromosomal DNA sequence expressed in testicular tissue. Nucleic Acids Res 1987; 15: 8713–8724. 37. Manz E, Schnieders F, Brechlin AM, Schmidtke J. TSPY-related sequences represent a microheterogeneous gene family organized as constitutive elements in DYZ5 tandem repeat units on the human Y chromosome. Genomics 1993; 17: 726–731. 38. Schnieders F, Dork T, Arnemann J, Vogel T, Werner M, Schmidtke J. Testis-specific protein, Y- encoded (TSPY) expression in testicular tissues. Hum Mol Genet 1996; 5: 1801–1807. 39. Lau Y, Chou P, Iezzoni J, Alonzo J, Komuves L. Expression of a candidate gene for the gonadoblastoma locus in gonadoblastoma and testicular seminoma. Cytogenet Cell Genet 2000; 91: 160–164. 40. Arnemann J, Jakubiczka S, Thuring S, Schmidtke J. Cloning and sequence analysis of a human Y-chromosome-derived, testicular cDNA, TSPY. Genomics 1991; 11: 108–114. 41. Zhang JS, Yang-Feng TL, Muller U, Mohandas TK, de Jong PJ, Lau YF. Molecular isolation and characterization of an expressed gene from the human Y chromosome. Hum Mol Genet 1992; 1: 717–726. 42. Shi SR, Key ME, Kalra KL. Antigen retrieval in formalin-fixed, paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991; 39: 741–748. 43. Castrillon DH, Quade BJ, Wang TY, Quigley C, Crum CP. The human VASA gene is specifically expressed in the germ cell lineage. Proc Natl Acad Sci U S A 2000; 97: 9585–9590. 44. Beckstead JH. Alkaline phosphatase histochemistry in human germ cell neoplasms. Am J Surg Pathol 1983; 7: 341–349. 45. Jacobsen GK, Norgaard-Pedersen B. Placental alkaline phos-phatase in testicular germ cell tumours and in carcinoma-in-situ of the testis. An immunohistochemical study. Acta Pathol Microbiol Immunol Scand A 1984; 92: 323–329. 46. Rajpert-De Meyts E, Skakkebæk NE. Expression of the c-kit protein product in carcinoma-in- situ and invasive testicular germ cell tumours. Int J Androl 1994; 17: 85–92. 47. Strohmeyer T, Reese D, Press M, Ackermann R, Hartmann M, Slamon D. Expression of the c- kit proto-oncogene and its ligand stem cell factor (SCF) in normal and malignant human testicular tissue. J Urol 1995; 153: 511–515. 48. Zeeman AM, Stoop H, Boter M, et al. VASA is a specific marker for both normal and malignant human germ cells. Lab Invest 2002; 82: 159–166. 49. Parkinson MC. Carcinoma in situ in boys with cryptorchidism: when can it be detected? Br J Urol 1994; 73: 431–435.

46 CARCINOMA IN SITU MARKERS IN FETAL GERM CELLS

50. Muller¨ H, Skakkebæk NE. Risk of testicular cancer in subfertile men: case–control study. Br Med J 1999; 318: 559–562. 51. Pinkerton JH, McKay DG, Adams EC, Hertig AT. Development of the human ovary — a study using histochemical techniques. Obstet Gynecol 1961; 18: 152–181. 52. Horie K, Fujita J, Takakura K, et al. The expression of c-kit protein in human adult and fetal tissues. Hum Reprod 1993; 8: 1955–1962.

A:IM>KÛ

8E@68???

BHEA4?B9#4G;B?B:LÒÚÙ ÓØß ÔÚÚ ÒÚÚÖ

E<878@4AABA86>8E 4AF&GBBC E4A> 4L8E 4EFG8AB>8@8L8E <8:B4FGE

Abstract Human germ cell tumours (GCTs) have long fascinated investigators for a number of reasons. Being pluripotential tumours, they can differentiate into both extra-embryonic and embryonic (somatic) tissues. However, it has never been shown convincingly that, in humans, these tumours are truly totipotent and can also give rise to the germ lineage, the third major differentiation lineage occurring early during embryonic life. Using a number of newly available, distinct, immunohistochemical markers, such as OCT3/4, VASA and TSPY, the occurrence of germ cells was investigated in a number of germ cell tumours. Development of germ cells was identified in three independent non-seminomas, including two pure yolk sac tumours and one mixed tumour composed of yolk sac tumour and immature teratoma. Our finding indicates a previously unknown totipotent potential of human GCTs and raises the question of whether, under certain culture conditions, primordial germ cells could be derived from human GCT cell lines.

50 GERM CELL DIFFERENTIATION IN NON-SEMINOMATOUS TUMORS Introduction Germ cell tumours of adolescents and adults (GCTs) are at the cross-roads of tumour and developmental biology. They are pluripotent tumours with multiple possible fates. Two subgroups can be distinguished, seminomatous and non- seminomatous GCTs (for review, see [1]), and Figure 1 shows their developmental potential schematically. Seminomas resemble early germ cells (primordial germ cells/gonocytes). They show limited capacity to differentiate into somatic or extra-embryonic tissues, although they can switch to a non-seminomatous phenotype [2]. Non-seminomas mimic early embryonic development. Embryonal carcinoma (EC) cells, the stem cells of non-seminomas, are highly similar to embryonic stem (ES) cells [3] and can give rise to embryonic endo-, meso- and ectoderm and/or differentiate into extra-embryonal yolk sac and trophoblast. As they have never been shown to give rise to the germ line in humans, they seem to fall one step short of totipotency.

Recent studies show that germ cells can be derived from mouse and human ES cells [4–6]. So far, it has been impossible to identify the development of germ cells in human GCTs. Recently, antibodies against several markers present in early germ cells of both sexes (OCT3/4, a marker for pluripotency, which is also expressed in ES cells, and VASA, a germ line-specific protein) or exclusively in male germ cells (TSPY, the testis-specific protein, Y-encoded), have been described [7–9]. Our search for germ cell development in non-seminomas was significantly supported by recent studies on the presence of these proteins during normal testicular development [10].

Materials and methods Research on human tumour samples has been performed according to the Code for Proper Secondary Use of Human Tissue in The Netherlands, as developed by the Dutch Federation of Medical Scientific Societies (FMWV) (version 2002), and has been approved by an institutional review board (MEC 02.981).

Formalin-fixed, paraffin-embedded tissue blocks from germ cell tumours that had been collected between 1991 and 2003, in close collaboration with urologists and pathologists in the south-western part of The Netherlands, were retrieved from our archive. In addition, tumour samples were obtained from patients treated according to protocols led by the Department of Hematology/Oncology, University of Tübingen. All cases were reviewed and diagnosed by J.W.O., according to the WHO classification.

51 CHAPTER 4

Figure 1. Developmental potential of normal and malignant germ cells. Normal development after fertilization is depicted in the boxed area on the left. Both embryonic (soma) and extra-embryonic tissues (trophoblast and yolk sac) develop from the zygote. Cells from the proximal region of the epiblast contribute to the primordial germ cell (PGC) pool, the allantois and the extra-embryonic mesoderm. The fate of epiblast cells becoming PGCs is not predetermined, but is specified by response to localized signals. After puberty, the derivatives of PGCs, known as spermatogonia, will eventually give rise to mature germ cells that can be fertilized and restart the life cycle (oocyte in females and spermatid in males). Due to as-yet unknown pathogenetic hits during intrauterine life, developmentally arrested germ cells can give rise to germ cell tumours. In the post- pubertal testis, this is thought always to involve an intermediate stage, termed carcinoma in situ/intratubular germ cell neoplasia unclassified (CIS/ITGCNU). Seminomas resemble early germ cells and CIS/ITGCNU, whereas embryonal carcinoma (EC) cells that originate after reprogramming of CIS/ITGCNU are highly similar to embryonic stem cells from the inner cell mass and can mimic early intrauterine development after fertilization (see boxed area in the middle). EC can differentiate into choriocarcinoma (CH), teratoma (TE) or yolk sac tumour (YST). Here we describe for the first time that tumour cells from non-seminomas containing yolk sac histology can reactivate totipotency and show properties of immature and more mature germ cells (see boxed area to the right)

52 GERM CELL DIFFERENTIATION IN NON-SEMINOMATOUS TUMORS Histochemical and immunohistochemical staining For immunohistochemistry, 3 μm sections were incubated with the primary antibody, followed by biotinylated secondary antibodies for 30 min and a biotinylated streptavidin horseradish peroxidase- or alkaline phosphatase coupled complex. The antibodies and conditions used are indicated in Table 1. All stains were counterstained with haematoxylin.

Table 1. Antibodies (source) and detection method used for immunohistochemistry

Secondary antibody Pre (1 : 200)(biotinylated) Antibody Company Code treatment Dilution Code Visualization

PLAP Cell Marque CMC203 HIAR* 1:200 Dako E0413 ABCplx-ap† c-KIT Dako A4502 HIAR 1:500 Dako E0431 ABCplx-ap OCT3/4 Santa Cruz sc-8629 HIAR 1:1000 Vector BA9500 ABCplx-hrp‡ VASA Provided by D Castrillon [43] HIAR 1:2000 Dako E0431 ABCplx-ap TSPY Provided by Y Lau [39] None 1:3000 Dako E0431 ABCplx-ap BMP4 Novocastra 3H2 None 1:100 Dako E0431 ABCplx-ap SCF Santa Cruz sc-1302 HIAR 1:400 Vector BA9500 ABCplx-ap CD30 Dako Ber-H2 HIAR 1:100 Dako E0413 ABCplx-hrp CD34 Dako Qbend/10 None 1:20 Dako E0413 ABCplx-hrp MS-363-S CD61 Immunotech SZ21,2116 HIAR 1:300 Dako E0413 ABCplx-hrp Glycophorin C Dako M0820 HIAR 1:100 Dako E0413 ABCplx-hrp Lysozyme Dako A0099 Pronase 1:900 Dako E0431 ABCplx-hrp Myeloperoxidase Dako A0398 HIAR 1:5000 Dako E0431 ABCplx-hrp

* Heat-induced antigen retrieval [42]. † ABCplx-ap, streptavidin-biotin- alkaline phosphatase complex. ‡ ABCplx-hrp streptavidin-biotin- horseradish peroxidase complex.

Double-staining was performed by using a combination of the same detection method but with different substrates: Fast Blue/Naphthol ASMX phosphate (F3378 and N500, Sigma, Steinheim, Germany) for blue staining and 3-amino-9- ethyl-carbazole (A5754 and D4254, Sigma, Steinheim, Germany)/H2O2 for red staining, without counterstaining. Endogenous peroxidase activity and/or endogenous biotin was blocked using 3% H2O2 (for 5 min) and/or a blocking kit for endogenous biotin (Vector Laboratories, Burlingame, CA, USA) to prevent background staining.

Results

Detection of cells with germ cell characteristics in non-seminomas Markers for immature germ cells and other factors involved in fetal germ cell differentiation, e.g. germ cell/placental-like alkaline phosphatases (PLAP), stem cell factor receptor (c-KIT) and glycogen, were studied in a series of 34 GCTs of pure non-seminomatous histology (YSTs, TEs and ECs): in other words, none of these tumours contained a seminomatous component, as judged by morphology

53 CHAPTER 4 and immunohistochemistry. In three cases, originating from the testis, the mediastinum and the pituitary gland, respectively, early germ cells were identified by morphology and marker expression (see Figure 2). In two cases, germ cells were found loosely distributed in clusters throughout the yolk sac component of the tumours (Figure 2A–E), whereas in the third case they were localized in tube-like structures (this tumour contained both yolk sac and immature teratoma components). The germ cells showed consistent staining for OCT3/4 (Figure 2C), PLAP (Figure 2D), c-KIT, TSPY (Figure 2E) and glycogen. CD30 was absent (Figure 2B insert), ruling out that these cells are EC cells rather than embryonic germ cells [11]. The presence of VASA (Figure 2F and lower insert), specific for late migratory and postmigratory germ cells [8], indicates that a number of these cells have progressed beyond the earliest stage of germ line commitment. While the majority of cells were positive for either OCT3/4 or VASA (Figure 2F), double staining revealed co-expression of both markers in only a minority of cells (lower insert), which was also the case for OCT3/4 and TSPY (Figure 2F, upper insert). The observation that the cells of interest are negative for OCT3/4 demonstrate that they are not seminoma cells.

Detection of cells with characteristics of early haematopoietic cells Staining for c-KIT, CD34, lysozyme, myeloperoxidase, Glycophorin C and CD61 showed the presence of cells with characteristics of early haematopoiesis. In two cases, nests of cells resembling angiogenetic clusters (so-called haemangioblasts) could be identified by the presence of c-KIT, CD34 (Figure 2G insert), a marker for pluripotent haematopoietic stem cells, and glycophorin C, a marker for red cells and their precursors (Figure 2H insert). A number of other factors for haematopoietic differentiation (ie lysozyme, myeloperoxidase and CD61) were occasionally present. However, the cells lacked the morphological characteristics of haematopoietic differentiation, such as blast formation (data not shown).

All three cases showed the presence of bone morphogenetic protein 4 (BMP4) in the YST component (Figure 2G), mainly in areas showing hepatoid histology. BMP4 was not restricted to the areas containing germ cells. Moreover, BMP4 was also observed in non-seminomas (ie yolk sac tumours) without the presence of germ cell lineage differentiation (data not shown). Therefore it cannot be considered as a specific marker. Stem cell factor (SCF), the ligand for c-KIT, was found in the yolk sac component, most often in close association with the developing germ cells (Figure 2H).

54 GERM CELL DIFFERENTIATION IN NON-SEMINOMATOUS TUMORS Discussion An extensive immunohistochemical analysis was performed to study the developmental potential and cell fate commitment of non-seminomas. We describe the novel finding of germ cell development in non-seminomas. In three cases, small groups of cells showing characteristics of fetal germ cell differentiation, ie loss of OCT3/4 and increased staining for VASA and TSPY, were found. The same pattern of OCT3/4 and VASA expression was recently described upon spontaneous differentiation of germ cells from human ES cells in vitro [6] and during normal germ cell development in humans [10]. Our findings indicate that germ cells can develop in non-seminomas, in particular YSTs showing expression of supporting factors, such as BMP4 and SCF. It is unknown whether the YST is the direct precursor of the germ cells or whether both lines co-develop from one common ancestor. This seems a rather semantic question; however, as all non-seminomas originate from a common stem cell, ie the EC cell (for review, see [1]; see also Figure 1).

Mammalian germ cells originate from the proximal region of the epiblast. In mouse experiments, ES cells can be stimulated to differentiate into germ cells and extra-embryonic mesoderm when grafted to the proximal epiblast [12]. This indicates that the potential to become a germ cell may not be restricted to predetermined progenitor cells but can be induced by extracellular factors [13]. The nature of these factors is largely unknown. Recent findings indicate that BMP4 might play a crucial role [4].

In two of the three cases, haematopoietic differentiation was found, underlining the close developmental relationship between germ cells and haematopoietic precursor cells. In fact, it has been hypothesized that primordial germ cells could be haematopoiesis-initiating cells [14], a process in which BMP4 could be an important signal [4,15]. Clinically, the relationship between mediastinal GCTs with yolk sac histology and acute myeloid leukemias is well established [16,17]. The haematopoietic precursor cells in these tumours are likely the origin of the leukaemias [18]. The clonal origin of mediastinal GCTs and haematological malignancies indicate a common malignant stem cell [19].

Failure to demonstrate germ cell development in non-seminomas so far might have been due to a number of reasons. It is probably a rare event (so far found in 3/34 non-seminomas), possibly due to the absence of crucial factors.

Alternatively, development of germ cells may be hampered by the fact that these cells are highly susceptible to apoptotic stimuli, leading to early death. Most important, however, could be the fact that so far the markers to identify germ lineage differentiation were lacking. Only the use of a combination of the

55 CHAPTER 4 markers OCT3/4, VASA and TSPY has enabled us to identify germ cell development in non-seminomas. It is highly unlikely that the presence of these markers is due to aberrant expression, because of the consistency in loss of OCT3/4 and increase in VASA, as found during normal germ cell development. In addition, GCTs are known to show expression of various markers in accordance with their physiological pattern, dependent on their specific lineage of differentiation (for review, see [1]).

The fact that non-seminoma cells can differentiate into germ cells is exciting and should be exploited in vitro using pluripotent GCT cell lines, under culture conditions that enhance formation of germ cells [4,6,20]. The resulting germ cells could be traced using a specific reporter system [4,5] and studied for their expression profile [6]. Deriving germ cells from human GCTs would be more than a mere technical exercise. Even though the cells of origin show neoplastic properties (eg they are aneuploid), the system would clearly reflect many aspects of normal germ cell development. It could therefore provide a readily accessible tool for the investigation of mechanisms of human germ cell development, yet avoiding the ethical issues and legal restrictions associated with the use of normal primordial germ cells derived from human embryos or human ES cells.

Figure 2. Representative illustrations of the various immunohistochemical markers identified in the germ cell tumours showing germ cell differentiation. (A) Testicular yolk sac tumour, haematoxylin and eosin staining (magnification ×10; insert ×40). (B) Same case: Ș-fetoprotein (AFP) staining is seen as a red cytoplasmic signal. Note the stronger staining intensity in more solid yolk sac tumour areas, showing a hepatoid pattern (magnification ×10). Insert is staining for CD30, being negative (magnification 20). (C) Same case: OCT3/4 (brown nuclear signal)-positive cells are scattered throughout the tumour (magnification ×10; insert ×40). (D) Same case: single cells show a red positive cytoplasmatic staining reaction for placental alkaline phosphatase (PLAP) (magnification ×10; insert ×40). (E) Same case: single cells show a red positive nuclear and cytoplasmic staining reaction for testis-specific protein on the Y chromosome (TSPY). Note also the weaker staining intensity in more solid yolk sac tumour areas, showing a hepatoid pattern (magnification ×10; insert 40×). (F) Same case: double staining for OCT3/4 (red nuclear signal) and VASA (blue cytoplasmic signal). Immature germ cells show strong signal intensity for OCT3/4, but are negative for VASA. This germ cell-specific marker is exclusively seen in more mature germ cells. The VASA-positive cells are negative for OCT3/4, and are therefore not seminoma cells (magnification ×40). Upper insert: double-staining for OCT3/4 (red nuclear signal) and TSPY (blue cytoplasmic signal). Note the double positive cells, and the single OCT3/4-positive cells (magnification ×40). Lower insert: double-staining for OCT3/4 (red nuclear signal) and VASA (blue cytoplasmic signal), showing some double positive cells for OCT3/4 and VASA. (G) Mediastinal yolk sac tumour, bone morphogenetic protein 4 (BMP4) staining (red cytoplasmic signal). Note the stronger staining intensity in more solid yolk sac tumour areas (magnification ×10). Insert: same case staining for CD34 (magnification 20). (H) Testicular yolk sac tumour, same case as shown in A–F. Stem cell factor (SCF) is seen as a red cytoplasmic and membranous signal in the majority of yolk sac tumour cells (magnification ×10). Insert: same case, staining for glycophorin C (magnification ×40)

56 GERM CELL DIFFERENTIATION IN NON-SEMINOMATOUS TUMORS

Figure 2.

57 CHAPTER 4 Acknowledgements We thank the urologists and pathologists in the south-western part of The Netherlands for their cooperation in collecting the tumour samples. This work was supported by the Dutch Cancer Society (H.S., J.W.O., L.H.J.L.) and the Deutsche Krebshilfe, Dr Mildred Scheel Stiftung (F.H.)

References 1. Oosterhuis JW, Looijenga LHJ. Testicular germ cell tumors in a broader perspective. Nat Rev Cancer 2005;5:210–222. 2. Oosterhuis JW, Kersemaekers AM, Jacobsen GK, Timmer A, Steyerberg EW, van Weeren PC, et al. Morphology of testicular parenchyma adjacent to germ cell tumours; an interim report. APMIS 2003;111:32–42. 3. Donovan PJ, de Miguel MP. Turning germ cells into stem cells. Curr Opin Genet Dev 2003;13(5):463–471. 4. Toyooka Y, Tsunekawa N, Akasu R, Noce T. Embryonic stem cells can form germ cells in vitro. Proc Natl Acad Sci USA 2003;100:11 457–11 462. 5. Hubner K, Fuhrmann G, Christenson LK, Kehler J, Reinbold R, De La Fuente R, et al. Derivation of oocytes from mouse embryonic stem cells. Science 2003;300:1251–1256. 6. Clark AT, Bodnar MS, Fox M, Rodriquez RT, Abeyta MJ, Firpo MT, et al. Spontaneous. Differentiation of germ cells from human embryonic stem cells in vitro. Hum Mol Genet 2004;13:727–739. 7. Looijenga LHJ, Stoop H, De Leeuw PJC, De Gouveia Brazao CA, Gillis AJM, Van Roozendaal KEP, et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003;63:2244–2250. 8. Castrillon DH, Quade BJ, Wang TY, Quigley C, Crum CP. The human VASA gene is specifically expressed in the germ cell lineage. Proc Natl Acad Sci USA 2000;97:9585–9590. 9. Lau Y, Chou P, Iezzoni J, Alonzo J, Komuves L. Expression of a candidate gene for the gonadoblastoma locus in gonadoblastoma and testicular seminoma. Cytogenet Cell Genet 2000;91:160–164. 10. Honecker F, Stoop H, De Krijger R, Lau Y-FC, Castrillon DH, Bokemeyer C, et al. Pathobiological implications of the expression of markers of testicular carcinoma in situ by fetal germ cells. J Pathol 2004;203:849–857. 11. Latza U, Foss H-D, Durkop H, Eitelbach F, Dieckmann K-P, Loy V, et al. CD30 antigen in embryonal carcinoma and embryogenesis and release of the soluble molecule. Am J Pathol 1995;146:463–471. 12. Tam PP, Zhou SX. The allocation of epiblast cells to ectodermal and germ-line lineages is influenced by the position of the cells in the gastrulating mouse embryo. Dev Biol 1996;178:124–132. 13. Lawson KA, Dunn NR, Roelen BA, Zeinstra LM, Davis AM, Wright CV, et al. Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev 1999;13:424– 436. 14. Rich IN. Hemopoietic-initiating cells. J Perinat Med 1995;23:31–38. 15. Chadwick K, Wang L, Li L, Menendez P, Murdoch B, Rouleau A, et al. Cytokines and BMP-4 promote hematopoietic differentiation of human embryonic stem cells. Blood 2003;102:906–915. 16. Nichols CR, Roth BJ, Heerema N, Griep J, Tricot G. Hematologic neoplasia associated with primary mediastinal germ-cell tumors. N Engl J Med 1990;322:1425–1429.

58 GERM CELL DIFFERENTIATION IN NON-SEMINOMATOUS TUMORS

17. Hartmann JT, Nichols CR, Droz JP, Horwich A, Gerl A, Fossa SD, et al. The relative risk of second nongerminal malignancies in patients with extragonadal germ cell tumors. Cancer 2000;88:2629–2635. 18. Lee KC. Hematopoietic precursor cells within the yolk sac tumor component are the source of secondary hematopoietic malignancies in patients with mediastinal germ cell tumors. Cancer 1994;73:1535. 19. Woodruff K, Wang N, May W, Adrone E, Denny C, Feig SA. The clonal nature of mediastinal germ cell tumors and acute myelogenous leukemia. A case report and review of the literature. Cancer Genet Cytogenet 1995;79:25–31. 20. Geijsen N, Horoschak M, Kim K, Gribnau J, Eggan K, Daley GQ. Derivation of embryonic germ cells and male gametes from embry-onic stem cells. Nature 2004;427:148–154.

A:IM>KÜ

#"(ßÛ"'Ó Ô<78AG<9<8F68??FJ

4A68E%8F84E6;ÖÓ ÒÒÔÔ ÒÒßÚ ÒÚÚÓ

88A78EG BB<=8A:4 4AF&GBBC H58EG78 88HJ 4E?BF78BHI8<4E4M4B 7 CHAPTER 5

Abstract Human germ cell tumors (GCTs) may have variable histology and clinical behavior, depending on factors such as sex of the patient, age at clinical diagnosis, and anatomical site of the tumor. Some types of GCT, i.e., the seminomas/germinomas/dysgerminomas and embryonal carcinomas (the stem cell component of nonseminomas), have pluripotent potential, which is demonstrated by their capacity to differentiate into somatic and/or extraembryonic elements. Although embryonal carcinoma cells are intrinsically pluripotent, seminoma/germinoma/dysgerminoma cells, as well as their precursor carcinoma in situ/gonadoblastoma cells, have the phenotype of early germ cells that can be activated to pluripotency. The other types of GCT (teratomas and yolk sac tumors of infants and newborn, dermoid cyst of the ovary, and spermatocytic seminoma of elderly) are composed of (fully) differentiated tissues and lack the appearance of undifferentiated and pluripotent stem cells. OCT3/4, a transcription factor also known as OTF3 and POU5F1, is involved in regulation of pluripotency during normal development and is detectable in embryonic stem and germ cells. We analyzed the presence of POU5F1 in GCT and other tumor types using immunohistochemistry. The protein was consistently detected in carcinoma in situ/gonadoblastoma, seminomas/ germinoma/dysgerminoma, and embryonal carcinoma but not in the various types of differentiated nonseminomas. Multitumor tissue microarray analysis covering >100 different tumor categories and 3600 individual cancers verified that POU5F1 expression is specific for particular subtypes of GCT of adults. No protein was observed in GCT of newborn and infants, spermatocytic seminomas, and the various tumors of nongerm cell origin. In addition, no difference in staining pattern was found in chemosensitive and chemoresistant GCT of adults. These results indicate preservation of the link between POU5F1 and pluripotency, as reported during normal development, after malignant transformation. Therefore, POU5F1 immunohistochemistry is an informative diagnostic tool for pluripotent GCT and offers new insights into the histological heterogeneity of this cancer.

62 POU5F1 AND PLURIPOTENT GERM CELL TUMORS Introduction oct3/4, also known as otf3 or pou5f1, is a member of the POU family of transcription factors, which is expressed in pluripotent mouse and human embryonic stem and germ cells, including PGCs (1– 6). Expression of this gene is downregulated during differentiation (7). Furthermore, knocking out the pou5f1 gene in mice causes early lethality because of lack of inner cell mass formation (8) because pou5f1 is critical for self-renewal of embryonic stem cells (9). Interestingly, pou5f1 has been linked to the capacity of proper outgrowth of somatic cell clones (10). During human development, expression of POU5F1 is found at least until the blastocyst stage (11) in which it is involved in gene expression regulation. The protein activates transcription via octamer motifs located distally or proximally from transcriptional start sites (12). POU5F1 binding sites have been identified in various genes, including fibroblast growth factor 4 and the 1.5-kb alternative promoter of the platelet-derived growth factor Ș receptor (13). The data indicate that pou5f1/POU5F1 functions as a master switch in differentiation by regulating cells that have, or can develop, pluripotent potential.

We have previously demonstrated that POU5F1 transcripts are found in a specific set of human testicular GCT of adolescents and young adults (TGCT): the seminomas and embryonal carcinomas (14). In addition, the precursor lesions of TGCT, known as CIS (15), also express POU5F1 (14). These lesions are composed of cells that are considered to be the malignant counterpart of an embryonic germ cell, most likely a PGC (15–17). Interestingly, these cell types are in principle pluripotent or even multipotent (Ref. 18, for review). In contrast, no expression was found in the differentiated components of nonseminomas, i.e., teratomas, yolk sac tumors, and choriocarcinomas (14). Indeed, expression of POU5F1 has been reported in embryonal carcinoma cells lines, and downregulation of expression is found upon differentiation (4, 13).

In contrast to our finding of a specific expression pattern of POU5F1 in TGCT, expression of this gene has recently been reported in nonmalignant adult human tissues (19), as well as in a number of carcinoma cell lines (20). This latter finding was interpreted as the result of aberrant reactivation of embryonic genes during the process of malignant transformation. However, the conclusion that POU5F1 is expressed in these cells was based solely on reverse transcription PCR results, which can be misleading because of the presence of multiple POU5F1 pseudogenes (Refs. 14, 19, 21, own unpublished observations). We are not aware of any previous studies reporting analysis of POU5F1 protein expression in normal and malignant human tissues to clarify whether the mRNA is translated to functional POU5F1 protein. Therefore, an extensive immunohistochemical screening for POU5F1 protein expression was done in various types of GCT at 63 CHAPTER 5 different sites and in a set of >3600 tumors of >100 different types using multitumor tissue microarrays. POU5F1 immunoreactivity was detected only in cells of CIS/gonadoblastoma, seminoma/ germinoma/dysgerminoma, and embryonal carcinoma. These results convincingly demonstrate that the presence of POU5F1 protein is related to pluripotent capacity of human GCT and that reactivation of its expression is not a common mechanism in cancer. Conclusively, POU5F1 is a distinctive immunohistochemical marker to identify tumor cells resembling embryonic/primordial germ and embryonic stem cells.

Materials and methods

Sample Handling and Characterization The (T)GCT not included in the tissue arrays were collected in the southwestern part of the Netherlands in collaboration with urologists and pathologists. Representative parts of the tumor (and adjacent tissue, if available) were snap frozen in liquid nitrogen and were fixed in 10% formalin for paraffin embedding. Tumors were diagnosed according to the WHO classification (22), as previously described (23), supported by immunohistochemistry using antibodies directed against germ cell/ placental alkaline phosphatase, Ș-foeto protein, human chorionic gonadotropin, the stem cell factor receptor c-KIT, and cytokeratin (CAM5.2). The testicular tumors included 35 seminomas, 50 nonseminomas [with 14 embryonal carcinoma components, 21 teratoma components (6 mature, 5 immature, and 10 mixed), 18 yolk sac tumor components, and 5 choriocarcinoma components], and 10 spermatocytic seminomas. In addition, CIS containing testicular parenchyma (n = 16, including both adjacent to seminoma and nonseminoma) and embryonic testes of different developmental stages, i.e., from 17 to 40 weeks and 28 weeks postpartum, were included. These latter samples have been reported before (24). Moreover, 3 gonadoblastomas, found in dysgenetic gonads, of which 2 also contained dysgerminoma, as well as 4 GCT of the midline of the brain of adults (including 1 germinoma, 1 embryonal carcinoma, and 2 mixed differentiated nonseminomas) were included. To investigate possible difference in presence of POU5F1 between chemotherapy- sensitive and resistant GCT, a series of 34 patients with known clinical course, including 12 high-risk patients that were relapse-free after high-dose chemotherapy and 22 refractory cases, were investigated. Part of this series has been reported before (25, 26). The clinical parameters are indicated in Table 1. Patients were considered refractory when progression or relapse occurred despite adequate initial and salvage treatment, including high-dose chemotherapy with autologous stem-cell transplantation.

64 POU5F1 AND PLURIPOTENT GERM CELL TUMORS

Table 1. Clinical parameters of the patients with chemosensitive and chemoresistant GCT

Sensitive Resistant

Nos. of cases 12 22 Median age in years (range) 28 (20–47) 29 (17–56) Histologya Seminoma 1 1 Nonseminoma 11 21 Stage at diagnosis (UICCb) I 0 3 II 0 7 III 12 12 Initial treatment after surgery Surveillance 0 2 Chemotherapy 12 20 Follow-up (months) Median (range) 51 (14–69) 39 (11–180) Relapse free survival (months) Median (range) NR 7.1 (0–150) Response to initial treatment Complete remission 8 5 Partial remission, marker - 4 8 Partial remission, marker + 0 3 Progressive disease 0 3 Unknown 0 3 No. of salvage regiments Median (range) 0 3 (1–9) a For the POU5F1 immunostaining all seminoma, embryonal carcinoma, and CIS components of both the sensitive and resistant tumors (3 and 2; 6 and 8; and 4 and 6, respectively) are positive, whereas all differentiated nonseminoma components are negative. b UICC, International Union Against Cancer; NR, not reached.

To expand the series of both GCT and non-GCT, three different multitumor tissue microarrays were investigated. One was generated at the Department of Pathology of the University of Basel (Basel, Switzerland), and included 3273 interpretable individual tumors of >100 different types, represented on seven different slides (Table 2 and Fig. 2). The second was generated at the Division of Pediatric Pathology of the Johns Hopkins Medical Institution (Baltimore, MD) and included 84 (T)GCT of newborn, infants, and adolescents (Table 3), represented on four different slides. Both arrays contained various positive and negative controls. The third array, generated at the Department of Pathology, Erasmus Medical Center (GE Rotterdam, the Netherlands), included 48 esophagus tumors and 100 prostate cancers (50 progressing and 50 nonprogressing), as well as benign prostatic hypertrophy (n = 18). The result of this latter array is included in Table 1.

65 CHAPTER 5 Immunohistochemistry Immunohistochemistry with anti-POU5F1 antibodies was performed on paraffin- embedded tissue sections of 3-μm thickness. The sections were incubated overnight at 4°C with a polyclonal goat anti-POU5F1 antibody (C20, sc-8629; Santa Cruz Biotechnology, Santa Cruz, CA), directed toward the COOH terminus of the protein, diluted 1:8000 (final concentration, 0.025 μg/ml). Subsequently, a biotinylated horse anti-goat secondary antibody was applied to the sections, and the bound antibody complex was visualized using the horseradish peroxidase avidin-biotin complex method. Double fluorescence staining was performed using the polyclonal goat-anti-POU5Fl antibody and a monoclonal mouse-anti-c- KIT antibody (final concentration 2 μg/ml; Neomarkers, Fremont, CA) on frozen tissue sections. c-KIT is a known marker for CIS and seminomas (27). Secondary antibodies were labeled with FITC and CY3 (Dako Diagnostics, Glostrup, Denmark, and Jackson Immunoresearch, West Grove, PA, with a final concentration of 5 μg/ml), respectively. The stainings were also performed separately. Every experiment was accompanied by appropriate positive and negative (without primary antibody) controls.

Cell Culture The human cell lines Tera2 (28), 2102Ep (29), and NCCIT (30), all nonseminoma derived, were cultured and split under conventional conditions (37°C, 5% CO2). Cell lines were cultured until 80% confluency, and differentiation was induced with retinoic acid as described previously (30). Immunohistochemistry was performed on cytospin preparations as described above.

Results

POU5F1 Immunohistochemistry on TGCT On the basis of mRNA analysis (14), we expected that POU5F1 protein is present in specific different histological subtypes of TGCT. We first analyzed the presence of POU5F1 by immunohistochemistry on a series of TGCT of various histological types. Representative images of representative stainings are shown in Fig. 1. All tumor cells of seminomas and embryonal carcinomas showed a nuclear staining (Fig. 1, A and B), whereas all nontumor cells were negative. A similar pattern of staining was found in the gonadoblastomas, dysgerminomas and germinomas, and embryonal carcinomas of the brain (see also below). In contrast, all teratomas, both mature and immature, choriocarcinomas, and yolk sac tumors were negative (Fig. 1, C–E), as were the differentiated nonseminomas of the brain (data not shown). The staining intensity in the positive cases varied between moderate and high. None of the spermatocytic seminomas showed a positive staining (Fig. 1F). We also tested a series of CIS containing testicular parenchyma

66 POU5F1 AND PLURIPOTENT GERM CELL TUMORS samples, both adjacent to seminoma and nonseminoma. All CIS cells, identified by a double staining for c-KIT, irrespective of the histology of the invasive tumor, were positive for POU5F1 (Fig. 1G and inset). In contrast, no protein could be detected in any stage of spermatogenesis (see below). This makes POU5F1 one of the most informative immunohistochemical markers to identify CIS cells that our laboratory has tested.

Table 2. Results of tissue array immunohistochemistry of 3439 individual tumors of >100 different types for POU5F1 The organ (in alphabetic order), tumor type, number of cases studied, and positive findings are indicated

Organ Tumor Subgroup n Positive

Adrenal gland Carcinoma Adeno- 19 0 Anus Carcinoma Squamous 3 0 Brain Gliobl. multif./astrocytoma/ 46/35 0 oligodendroglioma/meningioma/ 24/46 0 PNET/ependymoma/ 17/9 0 gangioneuroma/ 7 0 medulloblastoma/estioneuroblastoma/ 4/2 0 opticus glioma/cranioph.a 1/8 0 Carotid body Glomus tumor 11 0 Cervix Carcinoma Squamous/CIS/adeno 37/7/4 0 Colon Adenoma Tubular and villous 134 0 Carcinoma Adeno- 47 0 Endometrium Carcinoma Adeno- 61 0 Sarcoma Endometrial stromal sarcoma 3 0 Esophagus Carcinoma Squamous/adeno 21/116 0 Gall bladder Carcinoma Adeno- 23 0 Glomus tumor 11 0 Kidney Carcinoma Clear cell/papillary 50/45 1/0 Chromophobic 13 0 Oncocytoma 7 0 Larynx Carcinioma Squamous 46 0 Liver Carcinoma Hepatocellular 41 0 Lung Carcinoma Adeno-/squamous/ 48/50 0/1 Large cell/small cell 47/41 1/0 Lymph nodes Lymphoma Hodgkin’s 50 0 Non-Hodgkin’s high/low grade 69 0 MALT 42 0 Breast Carcinoma Lobular/ductal 42/46 0 Med./muc./tub. 27/23/26 0 Cribriform/apocrine 5/3 0 Phylloidees tumor 13 0 Mouth Carcinoma Squamous 57 0 Ovary Carcinoma Endometroid/serous mucinous 49/48/20 0 Brenner/mixed Müllerian tumor 9/6 0 Granulosa cell 7 0 Dysgerminoma 2 2 Gonadoblastoma 1 1 Pancreas Carcinoma Adeno- 47 0 Paraganglia Paraganglioma 10 0 Parathyroid Carcinoma Adeno- 20 0 Parotid gland Carcinoma Mixed 15 0 Acinic cell/mucoepithelial 5/6 0

67 CHAPTER 5

Table 2: continued

Organ Tumor Subgroup n Positive

Phaeochrom. 26 0 Penis Carcinoma Squamous 43 0 Prostate Carcinoma Adeno-, horm. (non-)refract. 200 0 PIN/BPH 24/18 0 Salivary gland Pleom. adenoma 49 0 Lymphoep. lesion 5 0 Carcinoma Adenoid cystic 43 0 Acinic cell 36 0 Adenolymph. 30 0 Skin Carcinoma Squamous/basal 81/37 0 Melanoma 44 0 Merkel cell tumor 6 0 Benign Hemangioblastoma 8 0 Nevus 38 0 Appendix tumors 25 0 (Small) intestine Carcinoma Adeno- 17 0 GIST 14 0 Stomach Carcinoma Intestinal/diffuse 44/23 0 Soft tissue Benign Lipoma/fibrous histiocytoma/Schwann/ 15/58/35 0 neurofibroma/leiomyoblastoma/ 31/7 0 giant cell tumors of tendon sheet/ 29/ 0 hemangioma 17 0 Intermediate Haemangiopericytoma 8 0 Dermatofibrosarcoma protuberans 2 0 Malignant Kaposi/liposarcoma 25/26 0 Angiomyolipo./malign. Schwann. 1/8 0 Rhabdomyosarcoma/synovial sarcoma 13/3 0 Angios./fibros./alv. soft part sarc. 6-4-2001 0 Epithelioid sarcoma 2 0 Testis Seminoma/nonseminoma 49/47 49/33b Adenomatoid tumor 10 0 Thyroid Carcinoma Foll./papill./med. 42/48/36 0 Thymus Thymoma 23 0 Urinary bladder Carcinoma TCC 101 0 Small cell/squamous/adeno- 5-3-2004 0 Sarcoma 7 0 Uterus Myoma 107 0 Vagina Carcinoma Squamous 5 0 Vulva Carcinoma Squamous 37 0 Various Carcinoid 48 0 Adenomatoid tumor 10 0 Mesothelioma 23 0 Controls c 30 0 a Gliob. multif., glioblastoma multiforme; PNET, primitive neuro-ectodermal tumor; cranioph, craniopharyngioma; CIS, carcinoma in situ; MALT, mucosa associated lymphoid tissue; med, medullar; muc, mucinous; tub, tubular; horm (non-)refract, hormonal (non)refractory; angiomyolipo., angiomyoliposarcoma; malign. Schwann., malignant Schwannoma; anglo., angiosarcoma; fibros., fibrosarcoma; alv. soft part sarc., alveolar soft part sarcoma; foll., follical. b If containing seminoma in case of combined tumors or embryonal carcinoma elements. c Included in all seven slides: skin (n ൌ 4); lymph node (n ൌ 2); heart (n ൌ 2); skeletal muscle (n ൌ 2); smooth muscle (n ൌ 2); myometrium (n = 2); thyroid gland (n = 2); liver (n = 2); kidney cortex (n = 2); prostate (n = 2); testis (n = 2); submand. gland (n = 2); proliferative endometrium (n = 2); gray matter cerebrum (n ൌ 2).

68 POU5F1 AND PLURIPOTENT GERM CELL TUMORS To investigate whether a difference in presence of POU5F1 exists between chemosensitive and chemoresistant GCT, immunohistochemistry was done on a well-characterized series of GCT with known clinical outcome (Refs. 25, 26 and Table 2). The results indicate that the staining pattern is determined by histology, i.e., positive in CIS, seminoma and nonseminoma, irrespective of sensitivity to chemotherapy.

Table 3. Results of the multitissue microarray immunohistochemistry on 84 germ cell tumors of predominantly newborn, infants, and adolescents for POU5F1 The histology and anatomical localization of the tumor, sex of the patient, number of studied cases (n), age at clinical presentation and number of positive findings are indicated.

Histology Sexa Anat. Local. n Age (yr) Positive

Teratoma Total: 25 0 Mature M Sacral 2 0/1 0 F “ 2 0/0 0 M Nervous system 1 6 0 F Gonad 2 15/34 0 M “ 1 0,3 0 Immature F Sacral 3 0/0/0 0 M “ 1 0,3 0 F Gonad 11 5/6/8/10/10/11/ 0 12/12/13/15/?b M “ 1 0,8 0 ? ? 1 ? 0 Yolk sac tumor Total: 33 F Sacral 6 0.2/1/1/2/2.6/5 0 M “ 1 1,5 0 ? “ 6 0/0.5/?/?/?/? 0 M Gonad 10 0.7/0.7/1.3/1.3/ 0 1.9/2/2/2.8/3/4 F “ 9 7/8/9/10/10/13/ 0 18/20/24 Gonadoblastoma/ D Gonad 1 ? 1 Dysgerminoma Dysgerminoma F Gonad 11 12/12/12/17/18/ 1 14/14/?/?/?/? Dysgerminoma/ F Gonad 1 12 1 Nonseminoma Seminoma M Gonad 4 27/30/41/? 4 Mixed GCTc M Gonad 4 27/28/29/? 4d Nonseminoma M Gonad 5 23/?/?/? 5e Controlsf M/F 30 0 a M, male; F, female; ?, unknown; D, dysgenetic gonad. b ?, unknown. c Germ cell tumor. d When containing either a seminoma of embryonal carcinoma component. e When containing an embryonal carcinoma component. f POC (n = 9); fetal kidney (n = 4), fetal ovary (n = 1), fetal uterus (n = 2), Wilms’ tumor (n = 2), hepatoblastoma (n = 3), embryonal neuroblastoma (n = 5), testis (n = 3), MMMT (n ൌ 1).

69 CHAPTER 5 Immunohistochemistry for POU5F1 in Normal Spermatogenesis and Embryonic Germ Cells. Although POU5F1 could be detected by immunohistochemistry in CIS cells and cells of seminoma and embryonal carcinoma (see above), no staining was observed in seminiferous tubules containing normal spermatogenesis (Fig. 1G), neither on frozen nor formalin-fixed, paraffin-embedded tissue sections. In contrast, embryonic germ cells in embryos and fetuses were occasionally positive. The highest number and most intensely stained germ cells were found in gonads between week 17 and 24 of gestational age (Fig. 1H and see below). Earlier stages of development were not available for investigation. Only sporadic positive cells were identified in the samples obtained from developmental stages weeks 28, 35 and 37, whereas no staining was found in samples of 38 and 40 weeks gestational age and after birth. Germ cells are present in these samples, as identified by staining for various markers, including c-KIT and VASA (24, 31). These data indicate that POU5F1 is present in germ cells at least from the 17th to 37th week of gestational age, after which the protein is no longer detectable by immunohistochemistry.

POU5F1 in Undifferentiated and Differentiated Cells of TGCT-derived Cell Lines In the nonseminomatous TGCT-derived cell line Tera2, expression of the POU5F1 is only observed in undifferentiated cells, whereas induction of differentiation of embryonal carcinoma cells with retinoic acid results in down-regulation of POU5F1 expression. The potential of Tera2 cells to differentiate has been described before (4, 14). Here, we confirm this finding on the protein level by immunohistochemistry (Fig. 1I, left and right panel). In contrast, both the cell lines NCCIT and 2102Ep showed POU5F1 expression and protein before and induction with retinoic acid differentiation (data not shown). This is in accordance to the finding that these cell lines, in contrast to Tera2, do not show complete differentiation after exposure to retinoic acid (28, 30).

POU5F1 Immunohistochemistry on Multitumor Tissue Microarrays To investigate whether POU5F1 protein can be detected in other tumors than GCT, we performed immunohistochemistry using the same antibody on three different multitumor tissue microarrays. One contained a large series of tumors of different origin (Table 1), one of the other contained specifically prostate and esophageal carcinomas (of which the results are also included in Table 1), and one contained mainly GCT of newborn and infants (Table 3). None of the control tissues, both adult and embryonal (with one exception, see below), included in the arrays showed positive staining for POU5F1. Of the 3439 tumors (from the first arrays), all testicular seminomas and nonseminomas (containing either a

70 POU5F1 AND PLURIPOTENT GERM CELL TUMORS seminoma or an embryonal carcinoma component), as well as two dysgerminomas and one gonadoblastoma were positive, confirming the results presented above. Representative examples of the first arrays are shown in Fig. 2. In addition, one clear cell carcinoma of the kidney and two lung carcinomas (one squamous cell carcinoma and one undifferentiated large cell carcinoma) were positive, suggesting that POU5F1 protein can be found in a small percentage (1– 2%) of other tumors.

Fig. 1. Visualization of POU5F1 protein by immunohistochemistry on formalin-fixed, paraffin- embedded tissue sections of a seminoma (A-B); pure embryonal carcinoma (C); mixed embryonal carcinoma and teratoma (D); mixed embryonal carcinoma and choriocarcinoma (E); mixed embryonal carcinoma and yolk sac tumor (F); spermatocytic seminoma (G); testicular parenchyma containing seminiferous tubules with normal spermatogenesis and CIS cells. Inset shows a seminiferous tubules with CIS cells double stained for c-KIT (green) and POU5F1 (red) (H); an embryonic testis of 21 weeks of gestation (I); cytospins of Tera2 cells before (left panel) and after (right panel) exposure to retinoic acid, inducing differentiation, and resulting in loss of POU5F1.

The multitumor tissue microarray of the (T)GCT confirmed the specificity of the POU5F1 immunohistochemistry for the histological types CIS/gonadoblastoma, seminoma/dysgerminoma, and embryonal carcinoma, as described above (Table 3). None of the immature and mature teratomas and yolk sac tumors showed a positive staining. Again, embryonic germ cells, included as arrays in the array, were also positive (data not shown). These data establish POU5F1 as highly specific marker for normal and malignant embryonic germ cells and embryonic stem cells, both in precursor (CIS) lesions and invasive tumors.

71 CHAPTER 5 Discussion Human GCTs comprise a heterogeneous group of neoplasms, which are predominantly found along the midline of the body. This localization can be explained by the migration route of the PGC during embryogenesis from the yolk sac to the genital ridge (32). The histology and clinical behavior of the various types of GCT depends on the sex of the patient, the age at clinical diagnosis, the anatomical location, and histology of the tumor (Ref. 33 for review). Overall, four different entities can be distinguished: (a) the teratomas and yolk sac tumors of newborn and infants; (b) the seminomas/dysgerminomas/ germinomas and nonseminomas of the gonads (both ovary and testis, and dysgenetic gonad), anterior mediastinum, and midline of the brain; (c) the dermoid cyst (mature cystic teratoma) of the ovary; and (d) the spermatocytic seminoma of the elderly testis. From a biological point of view, the pluripotent tumors (entity b) are the most intriguing. Histologically, they are subdivided into seminomas (known as dysgerminoma when identified in the ovary and dysgenetic gonad, and germinoma when arising in extragonadal sites) and non-seminomas (Ref. 34 for review). The seminoma-like tumors are composed of cells with a similar morphology as CIS cells. The nonseminomas can contain different histological elements, including embryonal carcinoma as the stem cell component, teratoma (representing somatic differentiation), yolk sac tumor, and choriocarcinoma (representing extraembryonic differentiation). The occurrence of embryoid bodies in nonseminomas (22), as well as certain patterns of gene expression, including POU5F1 (4, 13, 14, 35–38), illustrates the similarities between developmental patterns in the embryo and non-seminomas. This study aimed at a further understanding of the nature of the pluripotent (T)GCT compared with the other types of GCT and tumors of nongerm cell origin.

Here, we convincingly show that testicular CIS, seminoma, and embryonal carcinoma cells are consistently positive for POU5F1, irrespective of chemosensitivity or chemoresistance. Gonadoblastomas and associated dysgerminomas, as well as germinomas and embryonal carcinomas of the midline of the brain, showed a similar pattern. In contrast, none of the differentiated nonseminomatous derivatives (teratoma, yolk sac tumor, and choriocarcinoma) stained positive, irrespective of anatomical localization. This suggests that loss of protein expression is because of down-regulation of gene expression upon differentiation/maturation. These data are in accordance with earlier findings in mice, indicating that pou5f1 is indeed a marker for embryonic stem cells and germ cells (1–3) and that expression is lost upon additional differentiation and maturation (7). The fact that we observed a homogeneously positive staining for POU5F1 in the embryonal carcinoma components of both pure and mixed nonseminomas (and representative cell lines) supports the

72 POU5F1 AND PLURIPOTENT GERM CELL TUMORS model that the encoded protein is crucial but not sufficient for pluripotency. This has also been implied by the recent finding that POU5F1 overexpression had no effect on HeLa cells (39). Apparently, particular auxiliary proteins are needed. In accordance with the histology-dependent pattern of POU5F1 staining, both mature and immature teratomas, as well as the yolk sac tumors of newborn and infants, are negative. This is also true for the spermatocytic seminomas. Both tumor types have a different pathogenesis from TGCT (22, 24, 40 – 47). The absence of POU5F1 in these tumors is in agreement with their inability to generate pluripotent stem cells.

Our findings within in vivo tumor samples and representative cell lines are concordant with previous findings in mice (8, 9), showing that pou5f1 is expressed in pluripotent cells and is down-regulated upon differentiation. Therefore, we conclude that POU5F1 is involved in regulation of differentiation of human cells in line with a report on human embryos (6). No POU5F1 was detectable in adult testicular samples with normal spermatogenesis by immunohistochemistry, which is consistent with findings in mice using fluorescent protein-tagged pou5f1 expression analysis (48). However, a positive staining by immunohistochemistry has been reported in mouse spermatogonia A (7). This difference might be because of sensitivity differences of the methods applied and/or differences in protein level. The finding of POU5F1 in human embryonic germ cells and the down-regulation of expression before spermatogonia A formation supports an embryonic initiation of the pathogenesis of TGCT (15, 16, 27, 49 –52).

We did not find POU5F1 protein in any of the differentiated components of nonseminomas or in the majority (>99.9%) of 3340 nongerm cell tumors. This indicates that reactivation of POU5F1 expression and translation, as previously suggested based on cell line data (20), is not a frequent finding in cancer. The positive tumors were one kidney carcinoma (of 50 clear cell carcinomas) and two lung carcinomas (1 of 50 squamous and 1 of 47 large cell carcinomas analyzed). We do not currently have any plausible biological explanation for POU5F1 expression in these tumor types. Nevertheless, rare occurrence of POU5F1 immunoreactivity in kidney and lung must be kept in mind when performing immunohistochemistry for POU5F1 to demonstrate a pluripotent GCT in these organs. We think that the discrepancy between our study and the previous report suggesting reactivation of POU5F1 occurring as part of the malignant transformation process (20) can be attributable to the different techniques used. The mRNA expression analyses by Monk and Holding (20) were performed using reverse transcription-PCR with primers that may amplify both POU5F1 and different intron-less POU5F1-related sequences (unpublished observations).

73 CHAPTER 5

Fig. 2. A. schematic representation of one of the multitumor tissue microarrays used for the immunohistochemical detection of POU5F1, of which higher magnification examples are indicated in B. including a testicular seminoma (left panel) and an ovarian dysgerminoma (right panel).

In conclusion, this study demonstrates the usefulness of POU5F1 immunohistochemistry to support that TGCT originate from PGC and have pluripotent potential. However, POU5F1 is also present in nullipotent embryonal carcinoma and thus does not predict whether an embryonal carcinoma is able to generate various lineages of differentiation. Overall, POU5F1 is a highly specific immunohistochemical marker to confirm the diagnosis of CIS/gonadoblastoma, seminoma/dysgerminoma, and embryonal carcinoma. In addition, these data support the hypothesis that (T)GCT are excellent models to study the molecular mechanisms of pluripotency and differentiation, and therefore, a deeper understanding of these mechanism might also reflect into the rapidly evolving field of stem cell biology. Conversely, advances in the rapid evolving field of stem cell biology regarding the control of differentiation might also provide innovative therapeutic alternatives in (T)GCT in the future.

Acknowledgments We thank Ronald de Krijger (Department of Pathology, Erasmus Medical Center) for sharing the embryonic tissue samples.

74 POU5F1 AND PLURIPOTENT GERM CELL TUMORS References 1. Schöler, H. R., Dressler, G. R., Balling, R., Rohdewohld, H., and Gruss, P. Oct-4: a germline- specific transcription factor mapping to the mouse t-complex. EMBO J., 9: 2185–2195, 1990. 2. Okamoto, K., Okazawa, H., Okuda, A., Sakai, M., Muramatsu, M., and Hamada, H. A novel octamer binding transcription factor is differentially expressed in mouse embryonic cells. Cell, 60: 461– 472, 1990. 3. Rosner, M. H., Vigano, M. A., Ozato, K., Timmons, P. M., Poirier, F., Rigby, P. W., and Staudt, L. M. A POU-domain transcription factor in early stem cells and germ cells of the mammalian embryo. Nature (Lond.), 345: 686 – 692, 1990. 4. Pera, M. F., and Herszfeld, D. Differentiation of human pluripotent teratocarcinoma stem cells induced by bone morphogenetic protein-2. Reprod. Fertil. Dev., 10: 551–555, 1998. 5. Goto, T., Adjaye, J., Rodeck, C. H., and Monk, M. Identification of genes expressed in human primordial germ cells at the time of entry of the female germ line into meiosis. Mol. Hum. Reprod., 5: 851– 860, 1999. 6. Hansis, C., Grifo, J. A., and Krey, L. C. Oct-4 expression in inner cell mass and trophectoderm of human blastocysts. Mol. Hum. Reprod., 6: 999 –1004, 2000. 7. Pesce, M., Wang, X., Wolgemuth, D. J., and Schöler, H. Differential expression of the Oct-4 transcription factor during mouse germ cell differentiation. Mech. Dev., 71: 89 –98, 1998. 8. Nichols, J., Zevnik, B., Anastassiadis, K., Niwa, H., Klewe-Nebenius, D., Chambers, I., Schöler, H., and Smith, A. Formation of pluripotent stem cells in the mammalian embryo depends in the POU transcription factor Oct4. Cell, 95: 379 –391, 1998. 9. Niwa, H., Miyazaki, J., and Smith, A. G. Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nat. Genet., 24: 372– 376, 2000. 10. Boiani, M., Eckardt, S., Schöler, H. R., and McLaughlin, K. J. Oct4 distribution and level in mouse clones: consequences for pluripotency. Genes Dev., 16: 1209 –1219, 2002. 11. Abdel-Rahman, B., Fiddler, M., Rappolee, D., and Pergament, E. Expression of transcription regulating genes in human preimplantation embryos. Hum. Reprod., 10: 2787–2792, 1995. 12. Lamb, K. A., and Rizzino, A. Effects of differentiation on the transcriptional regulation of the FGF-4 gene: critical roles played by a distal enhancer. Mol. Reprod. Dev., 51: 218 –224, 1998. 13. Kraft, H. J., Mosselman, S., Smits, H. A., Hohenstein, P., Piek, E., Chen, Q., Artzt, K., and Van Zoelen, E. J. J. Oct-4 regulates alternative platelet-derived growth factor Ș receptor gene promoter in human embryonal carcinoma cells. J. Biol. Chem., 271: 12873–12878, 1996. 14. Palumba, C., Van Roozendaal, K., Gillis, A. J. M., Van Gurp, R. J. H. L. M., De Munnik, H., Oosterhuis, J. W., Van Zoelen, E. J. J., and Looijenga, L. H. J. Expression of the PDGF Ș- receptor 1.5 kb transcript. OCT-4 and c-KIT in human normal and malignant tissues. Implications for early diagnosis of testicular germ cell tumors and understanding regulatory mechanisms. J. Pathol., 196: 467– 477, 2002. 15. Skakkebæk, N. E. Possible carcinoma-in-situ of the testis. Lancet, 2: 516 –517, 1972. 16. Møller, H. Decreased testicular cancer risk in men born in wartime. J. Natl. Cancer Inst. (Bethesda), 81: 1668 –1669, 1989. 17. Jørgensen, N., Giwercman, A., Møller, J., and Skakkebæk, N. E. Immunohistochem-ical markers of carcinoma in situ of the testis also expressed in normal infantile germ cells. Histopathology, 22: 373–378, 1993. 18. Oosterhuis, J. W., and Looijenga, L. H. J. The biology of human germ cell tumours: retrospective speculations and new prospectives. Eur. Urol., 23: 245–250, 1993. 19. Takeda, J., Seino, S., and Bell, G. I. Human Oct3 gene family: cDNA sequences, alternative splicing, gene organization, chromosomal location, and expression at low levels in adult tissues. Nucleic Acids Res., 20: 4613– 4620, 1992.

75 CHAPTER 5

20. Monk, M., and Holding, C. Human embryonic genes re-expressed in cancer cells. Oncogene, 20: 8085– 8091, 2001. 21. Wey, E., Lyons, G. E., and Schafer, B. W. A human POU domain gene, mPOU, is expressed in developing brain and specific adult tissues. Eur. J. Biochem., 220: 753–762, 1994. 22. Mostofi, F. K., and Sesterhenn, I. A. Pathology of germ cell tumors of testes. Prog. Clin. Biol. Res., 203: 1–34, 1985. 23. Oosterhuis, J. W., Castedo, S. M. M. J., De Jong, B., Cornelisse, C. J., Dam, A., Sleijfer, D. T., and Schraffordt Koops, H. Ploidy of primary germ cell tumors of the testis. Pathogenetic and clinical relevance. Lab. Investig., 60: 14 –20, 1989. 24. Stoop, H., Van Gurp, R. H. J. L. M., De Krijger, R., Geurts van Kessel, A., Koberle, B., Oosterhuis, J. W., and Looijenga, L. H. J. Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological impli-cations. Lab. Investig., 81: 919 – 928, 2001. 25. Kersemaekers, A. M. F., Mayer, F., Molier, M., Van Weeren, P. C., Oosterhuis, J. W., Bokemeyer, C., and Looijenga, L. H. J. Role of P53 and MDM2 in treatment response of human germ cell tumors. J. Clin. Oncol., 20: 1551–1561, 2002. 26. Mayer, F., Gillis, A. J. M., Dinjens, W., Oosterhuis, J. W., Bokemeyer, C., and Looijenga, L. H. J. Microosatellite instability of germ cell tumors is associated with resistance to systemic treatment. Cancer Res., 62: 2758 –2760, 2002. 27. Rajpert-De Meyts, E., and Skakkebæk, N. E. Expression of the c-kit protein product in carcinoma-in-situ and invasive testicular germ cell tumours. Int. J. Androl., 17: 85–92, 1994. 28. Andrews, P. W. Pluripotent embryonal carcinoma clones derived from the human teratoma cell line Tera-2: differentiation in vivo and in vitro. Lab. Investig., 50: 147–167, 1984. 29. Wang, N., Perkins, K. L., Bronson, D. L., and Fraley, E. E. Cytogenetic evidence for premeiotic transformation of human testicular cancers. Cancer Res., 41: 2135–2140, 1981. 30. Damjanov, I., Horvat, B., and Gibas, Z. Retinoic acid-induced differentiation of the developmentally pluripotent human germ cell tumor-derived cell line. NCCIT. Lab. Investig., 68: 220 –232, 1993. 31. Zeeman, A. M., Stoop, H., Boter, M., Gillis, A. J. M., Castrillon, D. H., Oosterhuis, J. W., and Looijenga, L. H. J. VASA is a specific marker for both normal and malignant human germ cells. Lab. Investig., 82: 159 –166, 2002. 32. Anderson, R., Copeland, T. K., Schöler, H., Heasman, J., and Wylie, C. The onset of germ cell migration in the mouse embryo. Mech. Dev., 91: 61– 68, 2000. 33. Oosterhuis, J. W., Looijenga, L. H., van Echten, J., and de Jong, B. Chromosomal constitution and developmental potential of human germ cell tumors and teratomas. Cancer Genet. Cytogenet., 95: 96 –102, 1997. 34. Mostofi, F. K., Sesterhenn, I. A., and Davis, C. J. J. Immunopathology of germ cell tumors of the testis. Semin. Diagn. Pathol., 4: 320 –341, 1987. 35. Andrews, P. W., Fenderson, B. A., and Hakomori, S. Human embryonal carcinoma cells and their differentiation in culture. Int. J. Androl., 10: 95–104, 1987. 36. Van Gurp, R. J. L. M., Oosterhuis, J. W., Kalscheuer, V., Mariman, E. C. M., and Looijenga, L. H. J. Human testicular germ cell tumors show biallelic expression of the H19 and IGF2 gene. J. Natl. Cancer Inst. (Bethesda), 86: 1070 –1075, 1994. 37. Andrews, P. W., Casper, J., Damjanov, I., Duggan-Keen, M., Giwercman, A., Hata, J-I., Von Keitz, A., Looijenga, L. H. J., Oosterhuis, J. W., Pera, M., Sawada, M., Schmoll, H-J., Skakkebæk, N. E., Van Putten, W., and Stern, P. A comparative analysis of cell surface antigens expressed by cell lines derived from human germ cell tumors. Int. J. Cancer, 66: 806 – 816, 1996. 38. Looijenga, L. H. J., Gillis, A. J. M., Van Gurp, R. J. H. L. M., Verkerk, A. J. M. H., and Oosterhuis, J. W. X inactivation in human testicular tumors. XIST expression and androgen receptor methylation status. Am. J. Pathol., 151: 581–590, 1997.

76 POU5F1 AND PLURIPOTENT GERM CELL TUMORS

39. Tomilin, A., Vorob’ev, V., Drosdowsky, M., and Seralini, G. E. Oct3/4-associating proteins from embryonal carcinoma and spermatogenic cells of mouse. Mol. Biol. Rep., 25: 103–109, 1998. 40. Dekker, I., Rozeboom, T., Delemarre, J., Dam, A., and Oosterhuis, J. W. Placental-like alkaline phosphatase and DNA flow cytometry in spermatocytic seminoma. Cancer (Phila.), 69: 993– 996, 1992. 41. Burke, A. P., and Mostofi, F. K. Spermatocytic seminoma. A clinicopathologic study of 79 cases. J. Urol. Path., 1: 21–32, 1993. 42. Cummings, O. W., Ulbright, T. M., Eble, J. N., and Roth, L. M. Spermatocytic seminoma: an immunohistochemical study. Hum. Pathol., 25: 54 –59, 1994. 43. Rosenberg, C., Mostert, M. C., Bakker Schut, T., Van de Pol, M., Van Echten-Arends, J., De Jong, B., Raap, T., Tanke, H., Oosterhuis, J. W., and Looijenga, L. H. J. Chromosomal constitution of human spermatocytic seminomas: comparative genomic hybridization supported by conventional and interphase cytogenetics. Genes Chromosome Cancer, 23: 286 –291, 1998. 44. Riopel, M. A., Spellerberg, A., Griffin, C. A., and Perlman, E. J. Genetic analysis of ovarian germ cell tumors by comparative genomic hybridization. Cancer Res., 58: 3105–3110, 1998. 45. Perlman, E. J., Hu, J., Ho, D., Cushing, B., Lauer, S., and Castleberry, R. P. Genetic analysis of childhood endodermal sinus tumors by comparative genomic hybridization. J. Pediatr. Hematol. Oncol., 22: 100 –105, 2000. 46. Schneider, D. T., Schuster, A. E., Fritsch, M. K., Hu, J., Olson, T., Lauer, S., Gobel, U., and Perlman, E. J. Multipoint imprinting analysis indicates a common precursor cell for gonadal and nongonadal pediatric germ cell tumors. Cancer Res., 61: 7268 –7276, 2001. 47. Schneider, D. T., Schuster, A. E., Fritsch, M. K., Calaminus, G., Gobel, U., Harms, D., Lauer, S., Olson, T., and Perlman, E. J. Genetic analysis of mediastinal nonseminomatous germ cell tumors in children and adolescents. Genes Chromosome Cancer, 34: 115–125, 2002. 48. Yoshimizu, T., Sugiyama, N., De Felice, M., Yeom, Y. I., Ohbo, K., Masuko, K., Obinata, M., Abe, K., Schöler, H. R., and Matsui, Y. Germline-specific expression of the Oct-4/green fluorescent protein (GFP) transgene in mice. Dev. Growth Differ., 41: 675– 684, 1999. 49. Skakkebæk, N. E., Berthelsen, J. G., Giwercman, A., and Møller, J. Carcinoma-in-situ of the testis: possible origin from gonocytes and precursor of all types of germ cell tumours except spermatocytoma. Int. J. Androl., 10: 19 –28, 1987. 50. Meyts, E. R., Jorgensen, N., Muller, J., and Skakkebæk, N. E. Prolonged expression of the c-kit receptor in germ cells of intersex fetal testes. J. Pathol., 178: 166 –169,1996. 51. Møller, H., and Skakkebæk, N. E. Testicular cancer and cryptorchidism in relation to prenatal factors: case-control studies in Denmark. Cancer Causes Control, 8: 904 – 912, 1997. 52. Jørgensen, N., Møller, J., Jaubert, F., Clausen, O. P., and Skakkebæk, N. E. Heter-ogeneity of gonadoblastoma germ cells: similarities with immature germ cells, spermatogonia and testicular carcinoma in situ cells. Histopathology, 30: 177–186, 1997.

A:IM>KÝ

<4:ABFG<6I4?H8B9"'Ó Ô9BECE8

BHEA4?B9#4G;B?B:LÒÚÖ ÒÔÒ ÒÔØ ÒÚÚß

8EB8A78BA: 4AF&GBBC 8EG%B;?8 ;E8?<998A BBFGI4AFF8E 8G4? ';8F84HG;BEF6BAGE<5HG87 8DH4??LGÚG; CHAPTER 6

Abstract Human testicular germ cell tumours of adolescents and adults (TGCTs), the seminomatous and non-seminomatous germ cell tumours, show morphological and biological similarities to normal embryonic development, presumably determined by their supposed cell of origin, the primordial germ cell/gonocyte. Based on this knowledge, OCT3/4, also known as POU5F1, was recently defined as a diagnostic marker for these tumour types. In the adult testis, positive immunohistochemistry for OCT3/4 is an absolute indicator for the presence of the TGCT precursor carcinoma in situ/intratubular germ cell neoplasia undifferentiated (CIS/ITGCNU), seminoma, and/or embryonal carcinoma. Several studies have confirmed this observation, using the same polyclonal antibody. The present study demonstrates the usefulness of OCT3/4 immunohistochemistry in a diagnostic setting of a consecutively collected series of more than 200 testicular tumours and over 80 testicular biopsies. Moreover, it is shown that a monoclonal antibody directed against OCT3/4 is as informative as the polyclonal antibody, both in immunohistochemistry and in western blot analysis. The antibodies are robust and applicable with different methods of pretreatment and storage of tissue. This allows routine application of this diagnostic marker.

80 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS

Introduction Testicular germ cell tumours of adolescents and adults (TGCTs) seminomas and non-seminomas are the most common malignancies in Caucasian men in the second to fourth decades of life [1]. These tumours can be treated effectively using surgery, irradiation, and chemotherapy, resulting in 5-and 10-year survivals in the range of 90–95%. Despite this good prognosis, a major concern is the annual increase in incidence of 2–5% [2].

Invasive TGCTs exhibit a wide spectrum of histological types, ranging from seminoma to various non-seminomatous elements: embryonal carcinoma, teratoma, choriocarcinoma, and yolk sac tumour [3]. Seminoma cells are similar to embryonic germ cells, most likely primordial germ cells (PGCs) or gonocytes. Embryonal carcinoma is the malignant counterpart of embryonic stem cells, in principle pluripotent, and thus able to differentiate into all three germ lineages. This histological spectrum mimics embryonal development to a certain extent (see ref 4 for a review).

Despite the palette of histological appearances of TGCTs, they all originate from a common precursor, known as carcinoma in situ (CIS) [5] or intratubular germ cell neoplasia unclassified (ITGCNU). Recent findings regarding the expression of embryonic traits, and the high expression of the cell cycle regulators that are typical of embryonic mitotic germ cells in CIS cells, support the hypothesis that these cells originate from PGCs/gonocytes (see ref 6 for a review). So far, all known markers of CIS cells are also found in immature germ cells [7]. This hampers the applicability of these markers at an early age, particularly if a delay or block of maturation of germ cells is expected, for example in the case of cryptorchidism or gonadal dysgenesis [8]. The presence of immature germ cells may result in overdiagnosis of CIS, and therefore possibly overtreatment.

To detect CIS by means of immunohistochemistry, the markers c-KIT [9] and PLAP [10] have been used for more than a decade. One of the most recent markers is OCT3/4, also known as POU5F1. This transcriptional regulator is expressed in undifferentiated, pluripotent cells including embryonic stem cells and germ cells, both mouse [11–13] and human [14,15]. Knock-out mice are non-viable, because they fail to form a pluripotent inner cell mass, but instead consist primarily of cells committed to the trophoblast lineage [16]. Inhibition of expression of this gene in mouse and human embryonic stem cells results in a similar type of differentiation induction to trophectoderm [17,18]. In contrast, recent data show that PGCs without OCT3/4 do not differentiate, but undergo apoptosis [19]. This indicates that the role of OCT3/4 in embryonic stem cells and PGCs is different. We showed that expression of OCT3/4 in TGCTs is highly similar to the pattern found during normal development. In fact, testicular parenchyma containing CIS, 81 CHAPTER 6 as well as seminoma and embryonal carcinoma cells, contained mRNA of this gene, while it was absent in differentiated non-seminomas [20]. Subsequently, we demonstrated, by using a polyclonal antibody, that OCT3/4 protein is detectable in the cell types also positive for the mRNAs [21]. We applied immunohistochemistry to a large series of solid tumours (organized in tissue microarrays, TMAs) and found that OCT3/4 is a specific marker for CIS, seminoma, and embryonal carcinoma. In addition, its expression was also found in gonadoblastoma (the CIS counterpart of dysgenetic gonads), dysgerminoma, and germinoma (the seminomatous tumours of the ovary and midline of the brain, respectively). No correlation between treatment response and OCT3/4 staining was observed. None of the non-germ cell tumours investigated expressed the protein, except for three isolated cases [21]. Subsequently, these findings were confirmed by different studies [8,22–27] (see Table 1). In addition, Baker and Oliva recently reviewed the value of diagnostic markers in ovarian tumours, in which again OCT3/4 was concluded to be suitable for identifying dysgerminomas and ovarian embryonal carcinomas [28].

Table 1. Chronological summary of published data on OCT3/4 in germ cell tumours at the mRNA and protein levels

Germ cell tumours

Pre-invasive Invasive Metastatic

Seminomatous Nonseminomatous

Reference Method SpA CIS GB Se GE DG SpSE EC YST MTE ITE ChC EC SE

Looijenga et al [21] IHC - + + + + + - + - - - - Gidekel et al [24] IHC + + + + Sperger et al [22] Microarray + + - Rajpert-De Meyts et al [8] IHC - + + + - + Jones et al [25] IHC + - + - - - - Almstrup et al [23] Microarray + + + Cheng et al [26] IHC + + - - - Cheng [34] IHC + + Jones et al [27] IHC +

All immunohistochemical studies were performed with a polyclonal goat anti-human OCT3/4 Antibody (C-20, sc 8629; Santa Cruz Biotechnology, Santa Cruz, CA, USA), except for Gidekel et al, where a self-made rabbit anti-mouse OCT3/4 antibody was used. IHC = immunohistochemistry; SpA = spermatogonium A; CIS = carcinoma in situ; GB = gonadoblastoma; Se = seminoma; GE = germinoma; DG = dysgerminoma; SpSE = spermatocytic seminoma; EC = embryonal carcinoma; YST = yolk sac tumour; MTE = mature teratoma; ITE = immature teratoma; ChC = choriocarcinoma.

Here we report on our diagnostic experience of using immunohistochemistry for OCT3/4 on a consecutive series of more than 200 testicular cancers. Moreover, the applicability for the diagnosis of CIS in testicular biopsies is shown. We demonstrate that, in addition to the polyclonal antibody, which has been used in various studies, a suitable monoclonal antibody is now available that is

82 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS informative for both immunohistochemistry and western blotting. The monoclonal antibody assures infinite availability of OCT3/4 for clinical pathology and research. The diagnostic value of OCT3/4 immunohistochemistry is exemplified by the presentation of two patients. One presented with an unusual tumour in the brain and the other with a highly necrotic tumour in the retroperitoneal region. The positive staining of the neoplasms for OCT3/4 led to their recognition as germ cell tumours and the patients were treated successfully.

Materials and methods

Sample handling and characterization Use of tissues for scientific purposes was approved by the institutional review board (MEC 02.981). The samples were used according to the ‘Code for proper secondary use of human tissue in The Netherlands’, as developed by the Dutch Federation of Medical Scientific Societies (FMWV) (version 2002).

A total of 223 testicular tumours were consecutively collected in the south- western part of The Netherlands, in collaboration with urologists and pathologists. Representative parts of the tumour (and adjacent tissue, if available) were snapfrozen in liquid nitrogen or fixed in 10% formalin overnight for paraffin wax embedding. Tumours were diagnosed according to the WHO classification, supported by immunohistochemistry using antibodies directed against germ cell/placental alkaline phosphatase (PLAP), Ș-fetoprotein (AFP), human chorionic gonadotropin (hCG), the stem cell factor receptor c-KIT, and cytokeratin 8 (CAM5.2). CD30 was used on a selected series. In addition, 81 open surgical biopsies from 56 adults with primary infertility or a history of TGCT were evaluated for Johnson score and the presence of CIS.

Immunohistochemistry Immunohistochemistry with anti-OCT3/4 antibodies was performed on paraffin wax-embedded tissue sections of 3 μm thickness. Endogenous peroxidase activity was inactivated by incubation in 3% H2O2 for 5 min. Antigen retrieval was carried out by heating sections in 0.01 M sodium citrate (pH 6.0) under high pressure up to 1.2 bar. Endogenous biotin was blocked using an avidin/biotin blocking kit (SP-2001; Vector Laboratories, Burlingame, CA, USA). The sections were incubated for 2 h at room temperature with a polyclonal goat anti-human OCT3/4 antibody (sc-8629; Santa Cruz Biotechnology, Santa Cruz, CA, USA) directed against the COOH terminus of the protein. This antibody is used in almost all studies of OCT3/4 and germ cell tumours (Table 1). The antibody was diluted 1 : 1000. Subsequently, a biotinylated horse anti-goat secondary antibody

83 CHAPTER 6 (BA-9500; Vector) diluted 1 : 200 was applied and bound antibody complex was visualized using the horseradish peroxidase (HRP) avidin–biotin complex method. Sections were counterstained with haematoxylin.

A monoclonal antibody against the N-terminus of human OCT3/4 (sc-5279; Santa Cruz Biotechnology) was compared with the polyclonal antibody using a 1 : 1000 dilution in the same assay as described above. As a secondary antibody, a biotinylated rabbit anti-mouse antibody was used (E0413; Dako, Glostrup, Denmark). Detection and counterstaining were as described above. To compare the staining pattern of the monoclonal and polyclonal anti-OCT3/4 antibodies, immunohistochemistry was performed on parallel paraffin wax sections. In addition, both antibodies were used at a dilution of 1 : 4000 on parallel, 4 μm thick and acetone-fixed, frozen tissue sections.

Initial OCT3/4 staining of the brain tumour from patient 2 was performed on limited paraffin wax-embedded material that was left after a large panel of markers was tested. Therefore, only a small fragment containing tumour cells was present. To improve this result, a negative control slide was reused for OCT3/4 immunohistochemistry using the monoclonal antibody.

Immunofluorescence Fluorescent immunohistochemistry was performed on acetone-fixed frozen tissue sections of 4 μm thickness and cytospins of cell lines. The monoclonal anti- OCT3/4 antibody was incubated 1 : 2000 for 1 h at room temperature. Subsequently, a secondary FITC-labeled goat anti-mouse antibody (1090-02; Southern Biotechnology Associates, Birmingham, AL, USA) was incubated for 30 min. Counterstaining was done with Vectashield (Vector)/DAPI (Sigma Chemical Co, St Louis, MO, USA).

Cell culture The human cell lines Tera1 [29], 2102Ep [30], NT2 [31], and NCCIT [32], all of which were derived from embryonal carcinomas, were grown in a 5% humidified incubator until 80% confluency was reached. Cells were incubated with trypsin– EDTA, harvested, and cytospins were prepared.

84 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS

Protein isolation Tissue from frozen sections containing normal testis, embryonal carcinoma, yolk sac tumour, mature teratoma and seminoma, and pellets from cell lines Tera1, 2102Ep, NT2, and NCCIT were used to extract protein. Samples were washed in cold PBS and centrifuged at 0 0C for 5 min at 1500 rpm. Lysis buffer (1× PBS, 1% Nonidet P40, 0.5% sodium deoxycholate, 0.1% SDS) including protease inhibitors (1 836 153 complete mini cocktail; Roche Diagnostics, Mannheim, Germany) was added. Samples were sonicated and centrifuged at 0 0C for 10 min at 14 000 rpm. Protein lysates were stored at - 80 0C.

Western blot Samples of 20 μg protein were run on a 10% SDS-polyacrylamide gel, blotted on a PVDF membrane for 1 h, and incubated with a primary antibody overnight at 4 0C (anti-Oct3/4 monoclonal and polyclonal anti-body both diluted 1 : 1000). HRP- labelled secondary antibody was incubated for 1 h at room temperature. After washing, ECL western blotting detection reagents (RPN2109; Amersham Biosciences, UK) were added for 5 min and the blots exposed. After scanning, the blots were stripped and reprobed with anti-ș-actin (1 : 5000) as a loading control.

Results

Immunohistochemistry To test the diagnostic use of OCT3/4, immunohistochemistry was performed using the polyclonal antibody on a series of consecutively collected cases. In total, 209 testicular germ cell tumours were investigated after radical orchidectomy: three spermatocytic seminomas and 206 TGCTs (110 pure seminomas, 50 pure non-seminomas, and 46 combined tumours, the latter containing both a seminoma and a non-seminoma component). Mean age and range in major tumour groups were according to the literature. None of the three spermatocytic seminomas or 14 non-germ cell tumours [five B-cell lymphomas, three Leydig cell tumours, three embryonal rhabdomyosarcomas, one sarcoma, one carcinoid, and one metastatic renal carcinoma (clear cell type)] stained for OCT3/4. In contrast, all seminoma and embryonal carcinoma components were positive. CIS cells in the adjacent parenchyma of seminomas, non-seminomas, and combined tumours were all consistently positive. This was also true for intratubular seminoma (n = 29) and embryonal carcinoma (n = 18). There was a sharp transition between positive and negative cells, coinciding with the border of CIS, seminoma, and embryonal carcinoma with adjacent tissue. The CIS and seminoma cells showed highly consistent and strong nuclear staining, whereas in embryonal carcinoma cells, diffuse cytoplasmic staining was seen in addition to

85 CHAPTER 6 the nuclear staining. Representative examples are shown in Figures 1A and 1B (seminoma in Figure 1A; seminoma and embryonal carcinoma in Figure 1B). Fluorescent detection confirmed this cytoplasmic staining to be absent in seminoma (Figure 1C) and CIS, and specific for embryonal carcinoma (Figure 1D). Because cytoplasmic staining was never seen in the negative controls, and was only found together with nuclear presence of the protein, we regard cytoplasmic staining to be an intrinsic characteristic of embryonal carcinoma. Based on morphological and immunohistochemical data, no CIS, seminoma or embryonal carcinoma cell was found to be negative. This pattern of staining is in agreement with our previous observations, in line with the published data of others (Table 1).

Besides the different tumour samples, four embryonal carcinoma-derived cell lines were investigated. All cell lines showed strong nuclear and faint cytoplasmic staining with both enzyme-based and fluorescent immunohistochemistry. Remarkably, all cell lines contained a number of cells without positive nuclear OCT3/4 staining. This heterogeneity was clearly greater in the pluripotent NT2 cell line (41% OCT3/4-negative) than in nullipotent cell lines [2102Ep (9%), NCCIT (18%), and Tera1 (14%)], possibly due to spontaneous differentiation (Figures 1E and 1F). OCT3/4-negative cells were predominantly located at the borders of cell aggregates, which is consistent with differentiation [33].

Besides the analysis of orchidectomy specimens (see above), a series of consecutively collected testicular biopsies was included. These were taken because of infertility or a history of TGCT. In total, 81 biopsies from 56 patients were included. Based on both morphology and immunohistochemistry for OCT3/4, CIS was found in open surgical biopsies in four patients: two had a prior TGCT of the contralateral testis and two presented with infertility. A clinically occult seminoma was diagnosed in the biopsy from another patient with prior TGCT. In addition, in two patients with a palpable testicular node, the excision biopsy showed seminoma in both cases.

86 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS

Figure 1. OCT3/4 immunohistochemistry using different embedding, fixation, and detection methods. (A) Nuclear staining (brown) of seminoma cells with a monoclonal antibody. (B) Embryonal carcinoma and seminoma incubated with a polyclonal antibody. Note the cytoplasmic staining in the angio-invasive embryonal carcinoma (arrows) compared with the clear cytoplasm of intratubular seminoma cells. Both slides are from formalin-fixed and paraffin wax-embedded tissue. Images C–F show FITC-labelled fluorescence with the monoclonal antibody as the primary step. Nuclear staining is in blue (DAPI) and OCT3/4 localization in green (FITC). (C) Frozen tissue from a seminoma. (D) Frozen tissue from an embryonal carcinoma. Note the difference in cytoplasmic staining pattern between seminoma and embryonal carcinoma cells. Cytospin preparations of (E) pluripotent embryonal carcinoma-derived cell line NT2 and (F) nullipotent 2102Ep. Note the OCT3/4-negative cells located at the periphery of the aggregates

87 CHAPTER 6 Comparison of the monoclonal and polyclonal antibodies Because of the, by definition, finite availability of a polyclonal antibody, and possible differences between batches, we studied the suitability of a commercially available monoclonal antibody directed against OCT3/4. In parallel, this monoclonal and the polyclonal antibody used to date (also see above) were compared using immunohistochemistry on both frozen and formalin-fixed, paraffin wax-embedded tissue, and western blot analysis. Both antibodies showed a similar pattern in both approaches, without background staining (data not shown). Western blot analysis was performed on total protein extracted from normal testis, seminoma, and the various histological types of non-seminoma, as well as the TGCT-derived cell lines (Figure 2A). These results are in accordance with the immunohistochemical data. The presence of the two bands around the expected size of 43 kD, corresponding to the longer OCT3/4 variant [13], is currently under investigation.

Robustness of the polyclonal and monoclonal antibodies for immunohistochemistry To check the stability of the polyclonal and monoclonal antibodies for immunohistochemistry, the following procedure was performed repeatedly (n = 4). Undiluted antibody was frozen at -200C, thawed, used for immunohistochemistry, and again stored at -200C. Staining results did not decrease after multiple cycles. Immunohistochemistry results for frozen acetone- fixed, paraffin wax-embedded, formalin-, formaldehyde- and Bouin (Figure 2B) - fixed material and pretreated by microwave or high pressure were all successful, and suitable for diagnostic purposes.

Demonstration of the clinical value of OCT3/4 immunohistochemistry The clinical value of immunohistochemistry for OCT3/4 in the diagnosis of solid tumours is demonstrated by the following two cases.

Case 1 A 37-year-old male presented with pain in the left flank for 1 year, related to an abdominal mass extending from the diaphragm to the aortic bifurcation with involvement of the left kidney and retroperitoneal lymphadenopathy. The patient showed small testicles on ultrasound. Serum LDH was strongly elevated up to 3568 U/l (normal: 0–449), while hCG levels were elevated to 29.5 IU/l (normal: 0– 1.9). Guided biopsies showed large areas of necrosis and fibrosis with fields and strands of closely packed neoplastic cells with intermingling of lymphocytes. The malignant cells had round nuclei with prominent nucleoli and intermediate to large amounts of clear cytoplasm.

88 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS

Figure 2. (A) Western blot analysis for OCT3/4 in different testicular germ cell tumours, derived cell lines, and normal testis. Note the specificity for seminoma and embryonal carcinoma. (B) OCT3/4 immunohistochemistry on Bouin-fixed tissue showing positive staining of CIS. (C, D) Retroperitoneal tumour with extensive necrosis from patient 1 showing weak cytoplasmic, focally membranous staining for c-KIT in red (C, positive cells indicated with arrows) and distinct brown nuclear staining for OCT3/4 (D). (E, F) Tumour localized in the brain of patient 2, with scattered cells with clear cytoplasm and prominent nucleoli (E) with positive staining for OCT3/4 (F)

The morphological differential diagnosis was ‘seminoma, Ewing/small blue round cell tumour and lymphoma’. Immunohistochemically, the tumour cells showed very weak cytoplasmic and focally membranous staining for c-KIT (Figure 2C), while scattered cells were positive for keratin 8 (not shown). No staining was found in tumour cells for PLAP, hCG, CD30, AFP, CD99, desmin, CD21, CD45, EMA,

89 CHAPTER 6 and TDT. The lymphocytes were positive for CD45. The tumour cells were consistently positive for OCT3/4 (Figure 2D), from which the diagnosis of seminoma was made and the patient was accordingly treated with chemotherapy. During standard chemotherapy, there was a rise in AFP levels, unusual for a seminoma. A non-seminomatous component was therefore suspected and the patient underwent retroperitoneal lymph node dissection (RPLND) and nephrectomy of the involved left kidney. In addition, his left testis, which showed microlithiasis on ultrasound, was removed. AFP levels normalized and no viable tumour tissue was found in the RPLND, kidney or testis specimens. After a rise in hCG level to 5.8 IU/l suggesting residual tumour, his right testis was removed; again, no tumour tissue was found and hCG normalized. Two and a half years after the initial diagnosis, the patient has no signs of recurrent disease.

Case 2 A 40-year-old female presented with visual disturbances, followed by slowly progressive left-sided motor and sensory disturbances. She also suffered from memory disturbances and was mildly spatially and temporally disorientated. A T2 weighted MRI scan showed a large area of high signal intensity in the periventricular white matter of the right more than the left occipital region and the corpus callosum, without enhancement on T1 weighted images after gadolinium administration and without mass effect. A tumour was considered unlikely and an acute demyelinating encephalomyelopathy was suspected. Repeated cerebrospinal fluid examination showed mononuclear pleiocytosis (10 cells per cubic mm; normal value: <3 cells), without tumour cells. Flow cytometry showed predominantly reactive T-lymphocytes, without evidence of tumour cells. A CT scan of the thorax and abdomen did not reveal any abnormalities. The CSF- serum IgG index was normal and oligoclonal bands were absent. No evidence of vasculitis or an inflammatory disease was found, and treatment with steroids was ineffective. A stereotactic biopsy of the intracerebral lesion showed blastlike neoplastic cells within a mononuclear infiltrate (Figure 2E). A wide panel of markers including those for a primary central nervous system lymphoma was applied: CD3, CD4, CD5, ALK-1, CD19, CD20, CD79a, CD45, CD30, S-100, MELAN-A, HB45, CD68, CD43, PLAP, hCG, AFP, and CD56 all showed no staining; there was some punctate keratin positivity and no final diagnosis was made. At first, her condition stabilized, but after a few months it deteriorated further and an MRI scan showed increasing focal areas of enhancement. Immunohistochemistry for OCT3/4 (Figure 2F) revealed the tumour to be a germinoma and she was treated with whole brain irradiation. Follow-up MR imaging showed disappearance of the enhancement and her condition improved, although focal deficits remained present. One and a half year after treatment, her condition is stable.

90 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS

Discussion After our initial publication [21] showing the sensitivity and specificity of OCT3/4 immunohistochemistry for the detection of a number of specific histological variants of human germ cell tumours, the results were largely confirmed in subsequent studies, both by immunohistochemistry with the same polyclonal antibody and at the mRNA level (Table 1). The first study included an extensive multi-tumour tissue microarray covering more than 100 different tumour categories and 3600 individual cancers. Strong nuclear expression was found in seminomatous tumours, irrespective of anatomical localization (like germinoma and dysgerminoma), and embryonal carcinoma, and their common precursor cell CIS (as well as gonadoblastoma), whereas the differentiated components of non- seminomas (yolk sac tumour, teratoma, and choriocarcinoma) were consistently negative. No correlation with chemotherapy sensitivity or resistance was found. Since then, it has been demonstrated that OCT3/4 immunohistochemistry is of value to diagnose metastatic seminoma and embryonal carcinoma [34]. The antibody applied in the different studies reported so far is of polyclonal origin, generated in goat. Therefore, availability of this antibody is finite. Here we have demonstrated for the first time that similar results can be obtained using a mouse monoclonal antibody, the availability of which is unlimited. These antibodies give the same results in immunohistochemistry and western blot analysis. Both antibodies appear to be robust and highly stable, and the immunohistochemical detection method is relatively insensitive to pretreatment and poor quality of the tissue samples.

That OCT3/4 is indeed of additional value in clinical pathology compared with the other available markers is demonstrated by the two cases reported in this paper. One of these patients had an unusual presentation of a germinoma (seminoma of the brain) and the other patient had a highly necrotic retroperitoneal seminoma, and both were successfully treated as such. OCT3/4 was the only positive marker found, even when applied on a slide used previously as a negative control. The negativity of the other markers could be due to the absence of these markers in the tissue under investigation and/or poor quality of the tissue.

The value of using immunohistochemistry for OCT3/4 to screen for CIS in testicular parenchyma biopsies taken in the context of an outpatient andrology clinic is also presented. The data again demonstrate the usefulness of OCT3/4 immunohistochemistry for the diagnosis of these pre-invasive cancer cells. It must be kept in mind that the presence of OCT3/4-positive cells in testicular parenchyma of patients below the age of 1 year does not justify the diagnosis of

91 CHAPTER 6 CIS [7], in particular when a germ cell maturation delay or block is expected. For this specific application, additional criteria have to be developed.

Recently, Tai et al [35] reported OCT3/4 protein expression in some cells in the basal layer of human skin, corresponding in their hypothesis to the presence of adult stem cells. To verify these results, we incubated a total of 20 skin sections from five different individuals with the same monoclonal OCT3/4 antibody; no positive cells were found and this discrepancy remains to be resolved.

The data on OCT3/4 are also highly interesting from the biological point of view. An oncogenic driving force of OCT3/4 in the pathogenesis of TGCTs has been suggested based on mouse studies [24]. This conclusion was drawn from results obtained with experiments using mouse embryonic stem cells, which have various expression levels of OCT3/4. These cells give rise to so-called teratocarcinomas, which are not the proper animal model for TGCTs (see ref 36 for a review). In addition, it has been shown that the effect of OCT3/4 on a PGC and an embryonic stem cell is indeed different [19]. The fact that, in TGCTs, OCT3/4 expression mimics the pattern found during normal development (ie no OCT3/4 expression is detected in the differentiated non-seminomatous components) supports the model that OCT3/4 is under the control of physiological expression machinery and is not an oncogene in its classical sense. Our findings and those of others are in line with OCT3/4 expression being influenced by neither ploidy nor parental genome composition [37]. In fact, TGCTs originate from an erased PGC/gonocyte [38] and are consistently aneuploid [39]. It remains to be established whether epigenetic modifications have a regulatory role in the expression of OCT3/4 in TGCTs, as indicated in mouse cells [40].

In conclusion, we have demonstrated the diagnostic value of OCT3/4 immunohistochemistry for CIS, seminoma, and embryonal carcinoma in a clinical set-ting. Use of the monoclonal antibody is preferable for this application.

Acknowledgements This study was financially supported by Translational Research Grant Erasmus MC, and the Dutch Cancer Society.

92 DIAGNOSTIC VALUE OF OCT3/4 FOR (PRE-)INVASIVE TGCTS

References 1. McGlynn KA, Devesa SS, Sigurdson AJ, et al. Trends in the incidence of testicular germ cell tumors in the United States. Cancer 2003; 97: 63–70. 2. Adami HO, Bergstrom R, Mohner M, et al. Testicular cancer in nine northern Europeancountries. Int J Cancer 1994; 59: 33–38. 3. Woodward PJ, Heidenreich A, Looijenga LHJ, et al. Testicular germ cell tumors. In World Health Organization Classification of Tumours. Pathology and Genetics of the Urinary System and Male Genital Organs, Eble JN, Sauter G, Epstein JI, Sesterhenn IA (eds), vol 7. IARC Press: Lyon, 2004; 217–278. 4. Looijenga LH, Oosterhuis JW. Pathobiology of testicular germ cell tumors: views and news. Anal Quant Cytol Histol 2002; 24: 263–279. 5. Skakkebæk NE. Possible carcinoma-in-situ of the testis. Lancet 1972; 2: 516–517. 6. Rajpert-De Meyts E, Bartkova J, Samson M, et al. The emerging phenotype of the testicular carcinoma in situ germ cell. APMIS 2003; 111: 267–278; discussion 278–269. 7. Honecker F, Stoop H, de Krijger RR, et al. Pathobiological implications of the expression of markers of testicular carcinoma in situ by fetal germ cells. J Pathol 2004; 203: 849–857. 8. Rajpert-De Meyts E, Hanstein R, Jorgensen N, et al. Developmen-tal expression of POU5F1 (OCT-3/4) in normal and dysgenetic human gonads. Hum Reprod 2004; 19: 1338–1344. 9. Rajpert-De Meyts E, Skakkebæk NE. Expression of the c-kit protein product in carcinoma-in- situ and invasive testicular germ cell tumours. Int J Androl 1994; 17: 85–92. 10. Giwercman A, Cantell L, Marks A. Placental-like alkaline phos-phatase as a marker of carcinoma-in-situ of the testis. Comparison with monoclonal antibodies M2A and 43-9F. APMIS 1991; 99: 586–594. 11. Okamoto K, Okazawa H, Okuda A, et al. A novel octamer binding transcription factor is differentially expressed in mouse embryonic cells. Cell 1990; 60: 461–472. 12. Scholer HR, Ruppert S, Suzuki N, et al. New type of POU domain in germ line-specific protein Oct-4. Nature 1990; 344: 435–439. 13. Rosner MH, Vigano MA, Ozato K, et al. A POU-domain transcription factor in early stem cells and germ cells of the mammalian embryo. Nature 1990; 345: 686–692. 14. Carpenter MK, Rosler E, Rao MS. Characterization and differen-tiation of human embryonic stem cells. Cloning Stem Cells 2003; 5: 79–88. 15. Goto T, Adjaye J, Rodeck CH, Monk M. Identification of genes expressed in human primordial germ cells at the time of entry of the female germ line into meiosis. Mol Hum Reprod 1999; 5: 851–860. 16. Nichols J, Zevnik B, Anastassiadis K, et al. Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4. Cell 1998; 95: 379–391. 17. Velkey JM, O’Shea KS. Oct4 RNA interference induces trophectoderm differentiation in mouse embryonic stem cells. Genesis 2003; 37: 18–24. 18. Hay DC, Sutherland L, Clark J, Burdon T. Oct-4 knockdown induces similar patterns of endoderm and trophoblast differenti-ation markers in human and mouse embryonic stem cells. Stem Cells 2004; 22: 225–235. 19. Kehler J, Tolkunova E, Koschorz B, et al. Oct4 is required for primordial germ cell survival. EMBO Rep 2004; 5: 1–6. 20. Palumbo C, van Roozendaal K, Gillis AJ, et al. Expression of the PDGF alpha-receptor 1.5 kb transcript, OCT-4, and c-KIT in human normal and malignant tissues. Implications for the early diagnosis of testicular germ cell tumours and for our understanding of regulatory mechanisms. J Pathol 2002; 196: 467–477. 21. Looijenga LH, Stoop H, de Leeuw HP, et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003; 63: 2244–2250.

93 CHAPTER 6

22. Sperger JM, Chen X, Draper JS, et al. Gene expression patterns in human embryonic stem cells and human pluripotent germ cell tumors. Proc Natl Acad Sci U S A 2003; 100: 13350– 13355. 23. Almstrup K, Hoei-Hansen CE, Wirkner U, et al. Embryonic stem cell-like features of testicular carcinoma in situ revealed by genome-wide gene expression profiling. Cancer Res 2004; 64: 4736–4743. 24. Gidekel S, Pizov G, Bergman Y, Pikarsky E. Oct-3/4 is a dose-dependent oncogenic fate determinant. Cancer Cell 2003; 4: 361–370. 25. Jones TD, Ulbright TM, Eble JN, et al. OCT4 staining in testicular tumors: a sensitive and specific marker for seminoma and embryonal carcinoma. Am J Surg Pathol 2004; 28: 935– 940. 26. Cheng L, Thomas A, Roth LM, et al. OCT4: a novel biomarker for dysgerminoma of the ovary. Am J Surg Pathol 2004; 28: 1341–1346. 27. Jones TD, Ulbright TM, Eble JN, Cheng L. OCT4: a sensitive and specific biomarker for intratubular germ cell neoplasia of the testis. Clin Cancer Res 2004; 10: 8544–8547. 28. Baker PM, Oliva E. Immunohistochemistry as a tool in the differential diagnosis of ovarian tumors: an update. Int J Gynecol Pathol 2005; 24: 39–55. 29. Fogh J. Cultivation, characterization, and identification of human tumor cells with emphasis on kidney, testis, and bladder tumors. Natl Cancer Inst Monogr 1978; 49: 5–9. 30. Wang N, Trend B, Bronson DL, Fraley EE. Nonrandom abnor-malities in chromosome 1 in human testicular cancers. Cancer Res 1980; 40: 796–802. 31. Andrews PW, Damjanov I, Simon D, et al. Pluripotent embryonal carcinoma clones derived from the human teratocarcinoma cell line Tera-2. Differentiation in vivo and in vitro. Lab Invest 1984; 50: 147–162. 32. Teshima S, Shimosato Y, Hirohashi S, et al. Four new human germ cell tumor cell lines. Lab Invest 1988; 59: 328–336. 33. Rudnicki MA, McBurney MW. Cell culture methods and induction of differentiation of embryonal carcinoma cell lines. In Teratocarcinomas and Embryonic Stem Cells. A Practical Approach. Robertson EJ (ed). IRL Press: Oxford, 1987; 19–49. 34. Cheng L. Establishing a germ cell origin for metastatic tumors using OCT4 immunohistochemistry. Cancer 2004; 101: 2006–2010. 35. Tai MH, Chang CC, Olson LK, Trosko JE. Oct4 expression in adult human stem cells: evidence in support of the stem cell theory of carcinogenesis. Carcinogenesis 2005; 26: 495–502. 36. Honecker F, Oosterhuis JW, Mayer F, et al. New insights into the pathology and molecular biology of human germ cell tumors. World J Urol 2004; 22: 15–24. 37. Liu L, Czerwiec E, Keefe DL. Effect of ploidy and parental genome composition on expression of Oct-4 protein in mouse embryos. Gene Expr Patterns 2004; 4: 433–441. 38. van Gurp RJ, Oosterhuis JW, Kalscheuer V, et al. Biallelic expression of the H19 and IGF2 genes in human testicular germ cell tumors. J Natl Cancer Inst 1994; 86: 1070–1075. 39. Oosterhuis JW, Castedo SM, de Jong B, et al. Ploidy of primary germ cell tumors of the testis. Pathogenetic and clinical relevance. Lab Invest 1989; 60: 14–21. 40. Hattori N, Nishino K, Ko YG, et al. Epigenetic control of mouse Oct-4 gene expression in embryonic stem cells and trophoblast stem cells. J Biol Chem 2004; 279: 17 063–17 069.

A:IM>KÞ

!BA

HEBC84A(EB?B:LßÔ ÛßÓ ÛÖÚ ÒÚÚÙ

!<8?FI4A4FG8E8A 4AF&GBBC 8EG%B;?8 %BA4?778*

Abstract Objective: Carcinoma in situ (CIS) is accepted as the precursor of the germ cell tumors of the adult testis. CIS cells are located within the seminiferous tubules and can be exfoliated into semen. We performed a study to detect CIS cells in semen using the highly specific immunohistochemical marker OCT3/4, potentially a method for non-invasive diagnosis.

Material and methods: In 2006, 41 men at risk for CIS of the testis were found eligible for this study. Indications for inclusions were a suspicious lesion on scrotal ultrasound investigation (n = 14), patients on surveillance after a history of a testicular tumor (n = 14), and 13 patients with bilateral testicular microlithiasis (TM).

Results: Three of the 13 men (23%) who underwent testicular biopsies for bilateral TM were histologically diagnosed with CIS (two bilateral), and their semen showed OCT3/4-positive cells in all cases. Twelve of the 14 patients (86%) with a solid mass were diagnosed with a TGCT with adjacent CIS in the parenchyma, and in 9 cases (75%) OCT3/4-positive cells were present in the semen. No OCT3/4-positive cells were found in patients with biopsies who did not show any evidence of malignancy.

Conclusion: This study demonstrates that OCT3/4-positive cells can be found in semen from the majority of patients with CIS. The observations indicate that there is probably a time window in which the CIS cells are exfoliated, which gives an opportunity for early detection of CIS cells in semen of men at risk for TGCT.

96 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN

Introduction Testicular germ cell tumors (TGCTs) of adults, seminomas and nonseminomas (also known as type II GCTs) [1], account for 1–2% of all malignancies in men. Although the disease is uncommon, an annual rise in incidence is observed in most Western countries [1–4]. In the Netherlands a 5% annual increase was found between 1990 and 2005 [5]. Carcinoma in situ (CIS) is accepted as the precursor of TGCT [6,7]. CIS cells originate from primordial stem cells or gonocytes that escape normal development at an early point during intrauterine development [8]. These CIS cells start to proliferate during puberty, presumably after a rise in sex hormone levels. Patients with CIS of the testis will develop testicular cancer within 5 yr in 50%, and probably all patients will develop testicular cancer ultimately [9].

Fig. 1. Postorchidectomy specimen showing distinct OCT3/4-positive cells (stained red) in the epididymal lumen.

CIS is frequently found in the adjacent parenchyma of TGCTs [10]. CIS cells are located inside the seminiferous tubules at the basal membrane, in the niche of the spermatogonia, but they can leave their original location and spill over into the lumen. In addition, CIS cells can show a pagetoid spread to neighboring tubules, the rete testis [11] and even epididymus ( Fig. 1).

In view of this, it is plausible that CIS cells are exfoliated in semen, as are spermatozoa. CIS is present in the testis long before a tumor develops and may be used for screening purposes in men at risk for TGCT, specifically in men with fertility problems (0.6%) and testicular microlithiasis (20%), men who were treated for TGCT, and men at risk for a contralateral tumor (5%) [12–14]. Type II TGCTs are highly sensitive to treatment, which is also true for CIS [9,15]. This sensitivity allows early local treatment, preserving hormonal function in most cases [16].

Ever since it was established that CIS is the precursor of TGCTs, efforts have been made to detect CIS cells in semen. The use of semen for detection of neoplastic cells in patients with testicular cancer was already suggested by Czaplicki in 1987

97 CHAPTER 7 [17] and Giwercman in 1988 [18]. Different methods such as fluorescent in situ hybridization [19], immunohistochemistry using Ap-2 gamma and PLAP [20–22], and immunohistochemistry with magnetic beads using the M2A antibody [23] proved to be unsuccessful or too laborious. In contrast to the markers mentioned above, OCT3/4 is a very robust nuclear marker that has proven to be an absolute and specific marker for CIS, seminoma, and embryonal carcinoma in testicular tissue, and is now used as a standard marker in diagnosing CIS and TGCTs [24– 26] ( Fig. 2A). Our goal of this pilot study was to develop a reliable staining method. Moreover, we tried to evaluate the use of the OCT3/4 marker for early detection of CIS cells in semen of patients with known risk factors for CIS or TGCT.

Patients and methods

Patient selection The Institutional Ethics Committee approved this study and all participants gave written informed consent (code MEC-2005-282).

In this pilot study we selected, on the basis of ultrasound evaluation, patients who were at risk of harboring CIS in their testicles. Between January 2006 and January 2007, a total of 42 patients and 15 controls were included (see Table 1 for group characteristics). Fourteen patients were diagnosed with a testicular mass or ultrasound-detected lesion suspected for malignancy; all except 1 underwent unilateral inguinal orchidectomy. This latter patient underwent an open testicular biopsy with frozen-section analysis. No malignancy was found on frozen section; therefore; no orchidectomy was performed. All orchidectomy patients were counseled for semen cryopreservation and were asked to donate a semen sample for this study. In 6 of the 14 orchidectomy patients, we were able to obtain a complete semen sample before orchidectomy. Eight of the 14 orchidectomy patients donated only a residual part of at least 0.3 ml after the majority was cryopreserved for future fertility treatments. Also, patients who were diagnosed with testicular microcalcifications (TMs) during analysis for male infertility, candidates for testicular biopsies were asked to participate. A total of 13 men with bilateral TMs and bilateral testicular biopsy were included and all donated a complete semen sample. Fourteen postorchidectomy patients were included. These patients were either diagnosed with a marker relapse (n = 12) during their active surveillance protocol or were informed about this study by their physician who requested their participation (n = 2). None of these patients had testicular abnormalities on clinical examination or ultrasound. Moreover, 15 normospermic patients from the andrology clinic with no known risk factors for TGCT were used as negative controls. All semen samples were produced by masturbation. 98 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN

Methods The testicular biopsies and the orchidectomy specimens were examined with the use of standard protocols for the detection of TGCT and CIS. The semen samples were allowed to liquefy after production and thereafter dissolved in 10% phosphate-buffered formalin for 1 h. In one patient who underwent a bilateral orchidectomy, the left testicle showed a spermatocele, whose content was aspirated and used for the detection of OCT3/4-positive cells. At least 0.3 ml semen was obtained if the patient participated. After fixation the samples were centrifuged for 20 min at 1600g; then the pellet was resuspended in phosphate- buffered saline and vibrated with the use of an automatic shaker to make a single-cell solution. Cytospins of this suspension were made on a strong adhesive microscope slide (Starfrost©) and were dried overnight. Immunohistochemistry with monoclonal anti-OCT3/4 (POU5F1, Santa Cruz sc-10, sc-5279) antibodies was performed on the formalin-fixed semen samples as described earlier [25]. After the first 15 patients, we modified our last staining step by converting from a 3-diaminobenzidine tetrahydrochloride (DAB, brown) to a 3-amino-9- ethylcarbazole (AEC, red) reaction to increase the contrast between the OCT3/4- positive cells and the spermatozoa, and to lower the background staining. No differences in cell morphology were seen between these techniques. As a positive control, microscope slides were used with OCT3/4-positive cells from an established TGCT cell line (NT2) [27]. The microscope slides were blinded and evaluated separately by two individuals and were scored positive if a distinct nuclear staining was seen in large cells with large nuclei and clearly recognizable cytoplasm. Doubtful staining was scored negative.

Results Three of the 13 men (23%) who underwent testicular biopsies for TM were diagnosed with CIS, which in 2 of the cases was bilateral. The 14 orchidectomy specimens showed a seminoma in 4 patients (of which 1 was bilateral), nonseminoma in 4 patients, and a combined tumor (containing both a seminoma and a nonseminoma component in a single tumor) in 2 patients. In 2 testicles CIS was found next to a burned-out tumor. In 1 patient who had a suspicious intratesticular hypoechogenic lesion, an open testicular biopsy was performed and the specimen immediately stained by a direct alkaline phosphatase staining method [28]. Histopathological evaluation of this sample showed a Sertoli cell only (SCO) pattern. No CIS or malignant component was found and therefore no orchidectomy was performed. One orchidectomy specimen showed a Leydig cell tumor.

99 CHAPTER 7

Table 1 – Oct3/4 staining results, listed by diagnosis

Group No. Diagnosis Mean age at time OCT3/4-positive of diagnosis (yr) staining

Suspicion of TGCT 14 4 seminomaa 33,8 2/4 (50%) 4 nonseminomasa 23,7 4/4 (100%) 2 combined tumorsa 26,6 2/2 (100%) 1 Leydig cell tumor 26 0 2 testis-containing CIS 28,9 1/2 (50%) with a burned-out tumor 1 Sertoli cell only 28 0

Postorchidectomy 14 11 nonseminomas 26,3 0 3 combined tumors 24,8 0

Bilateral testicular 13 3 CIS (2 bilateral) 34,4 3 (100%) microcalcifications 10 no malignancy 34 0

Control group 15 Not available — 0 Total 56 15 testis-containing CIS 12/15 (80%)

TCGT, testicular germ cell tumor; CIS, carcinoma in situ. a: All TGCT tumors had CIS in the adjacent parenchyma.

Fig. 2. (A) Seminiferous tubule with carcinoma in situ stained with the OCT3/4 marker (brown). (B and C) Microscopic slides of semen showing spermatozoa and OCT3/4-positive cells from two patients diagnosed with a testicular germ cell tumor.

All orchidectomy specimens containing a TGCT showed CIS in the tissue adjacent to the tumor. In 12 of the 15 patients (80%) harboring CIS at the time of semen donation, OCT3/4-positive cells were found in the semen sample ( Table 1). Two of the 3 negative patients were diagnosed with a seminoma. OCT3/4-positive cells were seen with distinct CIS morphology as shown in Fig. 2B and C. The number of OCT3/ 4-positive cells ranged from 1 to >10 per sample. No correlation was found between the extensiveness of CIS in the testis or ejaculate volume and the number of exfoliated CIS cells. In patients with bilateral CIS who donated at least two total ejaculates, OCT3/4-positive cells could be found in most successive samples. Patient characteristics and a comparison between the OCT3/4 antibody staining of semen and the histopathological findings are shown in Table 1.

100 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN

In the patient with the bilateral seminoma only, OCT3/4-positive cells could be found in the aspirated spermatocele but not in the semen sample. No OCT3/4 positive cells were found in the patients who had no histologically proven abnormalities of the testis, nor did we find OCT3/4-positive cells in the semen samples from the postorchidectomy group. Mean age of the three patients in which no OCT3/4-positive cell were found in the semen, although proven TGCT was higher compared with the patients in which OCT3/4-positive cells were seen, was 35.8 versus 28.4 yr, respectively ( p = 0.067). Mean age in the seminoma patients with or without OCT3/4-positive cells was 26.8 versus 40.9 yr, respectively ( p = 0.12).

Discussion CIS is an asymptomatic condition that can be found in testicular biopsies in infertile patient with risk factors for TGCT or in patients with abnormal findings on ultrasound examination [9]. A testicular biopsy is still the gold standard in diagnosing CIS, with a false-negative percentage of 0.5%; recently it was even suggested to take a two-site biopsy to increase the diagnostic yield [29]. Although biopsy is a very sensitive diagnostic test, it is reported that CIS cells are not always randomly distributed, explaining false negative test-results in some cases [30]. Since the discovery of exfoliated CIS cells in semen in 1988 [18], a great deal of effort has been put in finding a method to detect these cells in patients at risk for TGCTs [23,31]. A non-invasive detection method would be helpful to a certain extent [25].

In this study we were able to detect OCT3/4-positive cells in semen from all CIS- bearing patients. This specific patient group may benefit from early detection. Although other studies as well as our own study have shown that CIS cells can be detected in semen, there are still a number of pitfalls. Malignant cells from the invasive tumor are not likely to be exfoliated owing to the predominant intact architecture of the seminiferous tubules. The primary location of the CIS cells beneath the tight junctions of the Sertoli cells may prevent spread in the first stage before proliferation has started. There is probably a time window between the early stage of CIS and the time the tumor obstructs the tubules in which the CIS cells are being exfoliated. Also the impact of the change in environment on cell morphology of CIS cells passing through the male genital tract may prevent proper detection. In the third place, with this technique, the natural surrounding in which the CIS cells are easily recognizable is lacking, making a positive identification more difficult and totally dependent on immunohistochemistry and cell morphology. In our study all patients showed recognizable CIS cells stained with the OCT3/4 marker ( Fig. 2B and C). The use of immunohistochemistry to detect the CIS cells, although it is highly specific,

101 CHAPTER 7 should be performed by using a standardized method. Expertise to per-form these staining techniques accurately to avoid false-negative results is, therefore, of major importance. Moreover, the specificity of the antibodies used must be confirmed, and a standardized method for immunohistochemical detection must be applied, including proper positive and negative controls. Hoei-Hansen et al [22] reported a diagnostic rate for the detection of CIS cells in semen from men diagnosed with CIS only of the testis, using AP2-gamma of 50% and 0% for OCT3/4. They were, however, not able to compare these techniques because they had not stained the same slides for AP2-gamma as well as OCT3/4. Also possible sub-optimal immunohistochemical detection of OCT3/4 might be related to the published limitation. These authors also described that three false-positive results were seminal fluid that showed borderline AP2-gamma–stained cells. A subsequently performed testicular biopsy failed to confirm this diagnosis. We did not encounter any false-positive results using the OCT3/4 marker. Although we did not have pathological examination of the controls, no abnormalities were seen on ultrasound or clinical investigation, making the change of an existing TGCT very small [32]. All three cases in which no OCT3/4 cells were found in the ejaculate showed severe fibrotic tubules potentially hampering exfoliation of CIS cells. The incidence of pagetoid spread of CIS cells in the rete testis differs between patients with seminoma and nonseminoma, with less pagetoid CIS involvement in the seminoma group [11]. Patients with a seminoma often have an extensive host response, less CIS, and more atrophic tubules [10]. This lymphocytic infiltration may even result in a complete eradication of CIS in seminoma patients [10,33]. Seminoma patients generally are older than nonseminoma patients, and the risk of any of these factors inhibiting exfoliation is increased. In our study semen from two patients (50%) with seminomas did not show any OCT3/4-positive cells, whereas in all patients with non-seminomas, OCT3/4-positive cells were seen. A similar tendency was seen in the study of Hoei-Hansen et al [22], in which they saw a significant difference in positive cells exfoliated between seminoma (17.4%) and nonseminoma patients (56.6%). Age (and histology of the tumor) might therefore play an important role in the exfoliation of CIS cells, with a higher chance of finding these cells in patients at a young age.

The purpose of this study was to try to detect CIS cells in semen especially to diagnose CIS before an overt tumor has developed. Although TGCT is a highly curable disease, the potential long-term side effects of the treatment protocols used are numerous and severe. Long-term side effects will occur in approximately 20-30% and consist mainly of nephrotoxicity, ototoxicity, neurotoxicity, and gonadal damage [34]. CIS can be treated with orchidectomy or local radiotherapy with a curability rate reaching 100%, thus preventing potential hazardous chemotherapy in case CIS is left untreated and a testicular

102 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN tumor develops. In this study, two patients had bilateral CIS and were advised to undergo bilateral radiation therapy. However, in case of infertility, couples may choose to have artificial reproductive techniques performed before radiation treatment is performed. The third patient with unilateral CIS underwent an inguinal orchidectomy to prevent the potential scatter radiation on the contralateral testicle in case of radiation therapy.

The combination of the peak-incidence of TGCTs, the long-term presence and exfoliation of CIS cells, and well-known risk factors may provide a setting in which screening could be useful. Also the relative high incidence (5%) of patients with TGCT occurring in the contralateral testis underscores the need for close follow-up and early detection in this specific group [35]. Patients visiting a fertility clinic have a 20-fold greater incidence of testicular germ cell cancer than fertile men, and CIS is found in approximately 0.6%; therefore, this might also be a population that could benefit from screening [36]. In addition, infertile males with bilateral TM, who have approximately a 20% chance of harboring CIS, might specifically profit from this technique [12]. Owing to the relative low incidence, the high number of patients at risk, and the overall good survival in case of TGCT, this test should be easy to use, cheap, highly specific, and sensitive, making it cost-effective. We are aware that the numbers in our study are relatively small; therefore, the results presented here are preliminary. Future research will focus on a more specific enrichment of the CIS cells, integration of an automatic screening tool for the detection of OCT3/4-positive cells, an increase in the numbers to substantiate our results, performance of a cost-benefit analysis, and, perhaps, the discovery of treatments that eradicate CIS while preserving remaining spermatogenesis.

Conflicts of interest The authors do not have any commercial relation-ships related to this article.

References [1] Oosterhuis JW, Looijenga LH. Testicular germ-cell tumours in a broader perspective. Nat Rev Cancer 2005;5:210–22. [2] Huyghe E, Matsuda T, Thonneau P. Increasing incidence of testicular cancer worldwide: a review. J Urol 2003;170:5– 11. [3] Maffezzini M. TC incidence increasing: spread the word. Eur Urol 2007;51:596–7. [4] Huyghe E, Plante P, Thonneau PF. Testicular cancer variations in time and space in Europe. Eur Urol 2007;51:621–8. [5] Post PN, Casparie MK, Kate FJW, ten en Oosterhuis JW. Epidemiologie van testistumoren in Nederland: accurate weergave in de PALGA-registratie. Nederlands tijdschrift voor Geneeskunde 2004;148:1150–4. [6] Hoei-Hansen CE, Rajpert-De Meyts E, Daugaard G, Skakkebæk NE. Carcinoma in situ testis, the progenitor of testicular germ cell tumours: a clinical review. Ann Oncol 2005;16:863–8.

103 CHAPTER 7

[7] Looijenga LH, Oosterhuis JW. Pathobiology of testicular germ cell tumors: views and news. Anal Quant Cytol Histol 2002;24:263–79. [8] Honecker F, Stoop H, de Krijger RR, et al. Pathobiological implications of the expression of markers of testicular carcinoma in situ by fetal germ cells. J Pathol 2004;203: 849–57. [9] Rorth M, Rajpert-De Meyts E, Andersson L, et al. Carcinoma in situ in the testis. Scand J Urol Nephrol Suppl 2000;205:166–86. [10] Oosterhuis JW, Kersemaekers AM, Jacobsen GK, et al. Morphology of testicular parenchyma adjacent to germ cell tumours. An interim report. Apmis 2003;111:32–40, discussion 1–2. [11] Perry A, Wiley EL, Albores-Saavedra J. Pagetoid spread of intratubular germ cell neoplasia into rete testis: a morphologic and histochemical study of 100 orchiectomy specimens with invasive germ cell tumors. Hum Pathol 1994;25:235–9. [12] de Gouveia Brazao CA, Pierik FH, Oosterhuis JW, et al. Bilateral testicular microlithiasis predicts the presence of the precursor of testicular germ cell tumors in subfertile men. J Urol 2004;171:158–60. [13] Dieckmann KP, Pichlmeier U. Clinical epidemiology of testicular germ cell tumors. World J Urol 2004;22:2–14. [14] Dieckmann KP, Skakkebæk NE. Carcinoma in situ of the testis: review of biological and clinical features. Int J Cancer 1999;83:815–22. [15] Dieckmann KP, Besserer A, Loy V. Low-dose radiation therapy for testicular intraepithelial neoplasia. J Cancer Res Clin Oncol 1993;119:355–9. [16] Giwercman A, von der Maase H, Berthelsen JG, et al. Localized irradiation of testes with carcinoma in situ: effects on Leydig cell function and eradication of malignant germ cells in 20 patients. J Clin Endocrinol Metab 1991;73:596–603. [17] Czaplicki M, Rojewska J, Pykalo R, Szymanska K. Detec-tion of testicular neoplasms by cytological examination of seminal fluid. J Urol 1987;138:787–8. [18] Giwercman A, Marks A, Skakkebæk NE. Carcinoma-in-situ germ-cells exfoliated from seminiferous epithelium into seminal fluid. Lancet 1988;1:530. [19] Meng FJ, Zhou Y, Giwercman A, et al. Fluorescence in situ hybridization analysis of chromosome 12 anomalies in semen cells from patients with carcinoma in situ of the testis. J Pathol 1998;186:235–9. [20] Hoei-Hansen CE, Rajpert-De Meyts E, Carlsen E, et al. A subfertile patient diagnosed with testicular carcinoma in situ by immunocytological staining for AP-2gamma in semen samples: case report. Hum Reprod 2005;20:579–82. [21] Brackenbury ET, Grigor KM, McIntyre MA, Howard GC, Hargreave TB. Negative testicular biopsy and asynchronous bilateral testicular germ cell tumour. Eur Urol 1994;25:79–81. [22] Hoei-Hansen CE, Carlsen E, Jorgensen N, et al. Towards a non-invasive method for early detection of testicular neoplasia in semen samples by identification of fetal germ cell- specific markers. Hum Reprod 2007;22:167–73. [23] Meng FJ, Zhou Y, Skakkebaek NE, Marks A, Giwercman A. Detection and enrichment of carcinoma-in-situ cells in semen by an immunomagnetic method using monoclonal antibody M2A. Int J Androl 1996;19:365–70. [24] Looijenga LH, Stoop H, de Leeuw HP, et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003;63:2244–50. [25] de Jong J, Stoop H, Dohle GR, et al. Diagnostic value of OCT3/4 for pre-invasive and invasive testicular germ cell tumours. J Pathol 2005;206:242–9. [26] de Jong J, Looijenga LH. Stem cell marker OCT3/4 in tumor biology and germ cell tumor diagnostics: history and future. Crit Rev Oncog 2006;12:171–203. [27] Andrews PW, Damjanov I, Simon D, et al. Pluripotent embryonal carcinoma clones derived from the human teratocarcinoma cell line Tera-2. Differentiation in vivo and in vitro. Lab Invest 1984;50:147–62.

104 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN

[28] Giwercman A, Cantell L, Marks A. Placental-like alkaline phosphatase as a marker of carcinoma-in-situ of the testis. Comparison with monoclonal antibodies M2A and 43-9F. Apmis 1991;99:586–94. [29] Dieckmann K-P, Kulejewski M, Pichlmeier U, Loy V. Diagnosis of contralateral testicular intraepithelial neoplasia (TIN) in patients with testicular germ cell cancer: systematic two- site biopsies are more sensitive than a single random biopsy. Eur Urol 2007;51:175–85. [30] van Casteren NJ, Boellaard WPA, Dohle GR, et al. Heterogeneous distribution of ITGCNU in an adult testis; con-sequences for biopsy-based diagnosis. Int J Surg Pathol 2008;16:21–4. [31] Giwercman A, Hopman AH, Ramaekers FC, Skakkebæk NE. Carcinoma in situ of the testis. Detection of malignant germ cells in seminal fluid by means of in situ hybridization. Am J Pathol 1990;136:497–502. [32] Giwercman A, Muller J, Skakkebæk NE. Prevalence of carcinoma in situ and other histopathological abnormalities in testes from 399 men who died suddenly and unexpectedly. J Urol 1991;145:77–80. [33] Schutte B, Holstein AF, Schirren C. Macrophages lysing seminoma cells in patients with carcinoma-in-situ (CIS) of the testis. Andrologia 1988;20:295–303. [34] Kollmannsberger C, Kuzcyk M, Mayer F, et al. Late toxicity following curative treatment of testicular cancer. Semin Surg Oncol 1999;17:275–81. [35] Daugaard G, Giwercman A, Skakkebæk NE. Should the other testis be biopsied? Semin Urol Oncol 1996;14:8–12. [36] Raman JD, Nobert CF, Goldstein M. Increased incidence of testicular cancer in men presenting with infertility and abnormal semen analysis. J Urol 2005;174:1819–22, discussion 22.

105 CHAPTER 7 Letter to the Editor

Re: Niels J. van Casteren, Hans Stoop, Gert R. Dohle, Ronald de Wit, J. Wolter Oosterhuis, Leendert H.J. Looijenga. Non-invasive Detection of Testicular Carcinoma In Situ in Semen Using OCT3/4. Eur Urol 2008;54:153–60 van Casteren et al set out to establish a method for detecting precursor cells of testicular germ cell cancer (CIS: carcinoma in situ, or TIN: testicular intraepithelial neoplasia) in semen with the aim of replacing testicular biopsy as the current diagnostic standard. By employing immunocytochemical staining with OCT3/4, they identified exfoliated TIN-cells in seminal fluid. This result is novel and exciting. Nonetheless, I have some concerns with this method.

1. If the patient examined possesses two testicles, then a positive finding in semen does not disclose the afflicted side because, clearly, semen is composed of excretions from both sides. Even shortly after unilateral orchidectomy, semen analysis will still reflect the pooled production of both testicles. As is known from post-vasectomy evaluations, the seminal vesicles can store exfoliated spermatozoa for months. With respect to searching for contralateral TIN, semen analysis will not be meaningful for several months.

2. We do not know the sensitivity of semen analysis for TIN. There appear to be unsolved technical problems because OCT 3/4 staining of semen cells had apparently failed in the hands of another experienced group [1].

3. The standard method for diagnosis of TIN rests, first, on morphological features and, second, on histochemical criteria, as found in tissue specimens featuring intact seminiferous tubules. Accordingly, conservation of specimens by proper fixation had been urged [2] and mechanical damage to specimens had been strongly cautioned [3].

Searching for exfoliated TIN-cells in semen represents a change of paradigm with respect to diagnostics of TIN because this new technology solely rests on immunocytochemical staining. Some of the distinct morphological features of TIN-cells as found in the tissue compound of seminiferous tubules are not detectable in single exfoliated cells in semen. In light of the problems many pathologists still have in diagnosing TIN in biopsies [4], one might speculate that sporadic TIN-cells in seminal fluid might easily escape detection. Generally, if a diagnostic technique is to replace current standards, then this new technique should involve clear-cut advantages over the standard; for example, higher accuracy, less costs, or better handling. Frankly, at this point, the advantages of semen analysis are still elusive.

106 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN

4. Searching for contralateral TIN in patients with testis cancer is the most frequent condition of TIN-diagnostics. Testicular biopsy is usually performed under general anaesthesia simultaneously with orchidectomy for the cancer. Although the potential for psychological distress imposed by such a biopsy has not been formally evaluated, it appears obvious that it is minimal. Nor has the emotional burden secondary to the quest to produce an ejaculate been formally evaluated. Yet conceivably, the latter procedure could involve an unpleasant task for a young man left with the problem of coping with cancer.

In all, van Casteren and coworkers present an appealing novel approach for detecting TIN. However, in light of a well-established current standard with high accuracy [2] and low surgical morbidity [5], open questions exist with the new technique regarding its clinical usefulness, methodological safety, diagnostic sensitivity, and patient inconvenience.

Conflicts of interest: The author has nothing to disclose.

References [1] C.E. Hoei-Hansen, E. Carlsen, N. Jorgensen, H. Leffers, N.E. Skakkebæk and E. Rajpert-De Meyts, Towards a non-invasive method for early detection of testicular neoplasia in semen samples by identification of fetal germ cell-specific markers. Hum Reprod, 22 (2007), pp. 167–173. [2] C.E. Hoei-Hansen, I.A. Olesen and N. Jorgensen, et al. Current approaches for detection of carcinoma in situ testis. Int J Androl, 30 (2007), pp. 398–405. [3] K.-P. Dieckmann and V. Loy, False-negative biopsies for the diagnosis of testicular intraepithelial neoplasia (TIN) an update. Eur Urol, 43 (2003), pp. 516–521. [4] A.M. Winstanley, G. Mikuz, F. Debruyne, C.C. Schulman and M.C. Parkinson, Handling and reporting of biopsy and surgical specimens of testicular cancer. Eur Urol, 45 (2004), pp. 564–573. [5] K.-P. Dieckmann, V. Heinemann, U. Frey and U. Pichlmeier, How harmful is contralateral testicular biopsy? An analysis of serial imaging studies and a prospective evaluation of surgical complications. Eur Urol, 48 (2005), pp. 662–672.

Klaus-Peter Dieckmann, Albertinen-Krankenhaus, Urologische Abteilung, Suentelstrasse 11a, D-22457 Hamburg, Germany

107 CHAPTER 7 Letter to the Editor

Reply to Klaus-Peter Dieckmann’s Letter to the Editor re: Niels J. van Casteren, Hans Stoop, Gert R. Dohle, Ronald de Wit, J. Wolter Oosterhuis, Leendert H.J. Looijenga. Non-invasive Detection of Testicular Carcinoma In Situ in Semen Using OCT3/4. Eur Urol 2008;54:153–60 We thank Dr Dieckmann for his valuable comments on our paper in which we described the observation that precursor cells of type II testicular germ cell tumours (TGCTs) can be found in semen, based on immunohistochemical detection of OCT3/4 (POU5F1) [1].

There are a number of issues to clarify. First, the primary objective of our study was to evaluate if carcinoma in situ/testicular intraepithelial neoplasia (CIS/TIN) cells could be detected in semen using OCT3/4 immunohistochemistry. This is a relevant question to answer because until now, this was not convincingly shown. The use of OCT3/4 is the optimal target for identification of CIS/TIN cells; because of its specificity, as discussed in the manuscript, OCT3/4 is the best marker for CIS/TIN cells in the adult testis as compared to other markers. If OCT3/4 is unable to detect CIS/TIN in semen, it is unlikely that another marker will.

However, there are additional issues to keep in mind. In general, success of immunohistochemistry depends on many factors, including the antibody and protocol used as well as fixation method and possible antigen retrieval. Considering these points, we believe that if in the study of Hoei-Hansen and colleagues the slides with Ap2-gamma positive cells were stained with OCT3/4, CIS cells would be found [2]. However, as we noted, because of their study design, no comparison can be made between Ap2-gamma and OCT3/4.

After succeeding in our primary objective, ie, feasibility to detect testicular tumour cells in semen, the question arises of how to implement this method in clinical practice. We are aware that the achievements made are only part of the first step and that many aspects must be clarified before implementation in clinical practice is achieved. Dieckmann discussed a number of those aspects. Indeed, an ejaculate is composed of germ cells and fluids from both testicles, and if OCT3/4-positive cells are found in an individual with two testes, the origin of the CIS/TIN cells is not known. In these cases, a surgical testicular biopsy is still required for CIS diagnosis. In addition, for testicular biopsies immunohistochemistry, preferentially using OCT3/ 4-specific antibodies, is mandatory to allow accurate diagnosis. Although generally CIS/TIN cells can be identified using

108 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN morphological criteria alone, they can be missed in testicular biopsies without immunohistochemistry.

Because of the low incidence of TGCT and the invasiveness of the procedure, a testicular biopsy for patients at risk for TGCT, such as those with infertility, cryptorchidism, testicular atrophy, and a unilateral TGCT, is not the preferred screening tool. In fact, country-specific differences are present, eg, a contralateral testicular biopsy in the case of a unilateral TGCT is obtained in every patient in Germany and Denmark, whereas such practice is uncommon in the Netherlands. However, we believe that in men with signs of testicular dysgenesis such as microlithiasis, small testicular volume, and an inhomogeneous ultrasound pattern, a diagnostic procedure to detect CIS/TIN should be considered. Detecting CIS cells in semen could be the ideal screening tool for men at risk for TGCTs.

An interesting and easily accessible group for CIS/ TIN screening are males who visit a fertility clinic. These patients always provide semen samples to assess fertility capacity, part of which can be used for screening for the presence of CIS/TIN. The extra burden lies only in explaining to the patient that he may have a higher risk for type II TGCTs, especially if signs of testicular dysgenesis exist, or in cases of testicular atrophy. This may also apply to men with a history of TGCT and in men with testicular microlitiasis found in scrotal ultrasound. If CIS cells are detected for validation, a surgical testicular biopsy is still required.

In conclusion, we believe that our approach might result in a future screening tool; however, various issues must be clarified before it can be implemented into clinical use. Therefore, prospective studies are needed to assess the sensitivity of the test. Our current study was a proof of concept, which will hopefully stimulate further development of an accurate and early diagnostic semen test that can be used for screening men at risk for TGCT.

Conflicts of interest: The authors have nothing to disclose.

References [1] Van Casteren NJ, Stoop H, Dohle GR, de Wit R, Oosterhuis JW, Looijenga LHJ. Non-invasive detection of testicular carcinoma in situ in semen using OCT3/4. Eur Urol 2008;54:153–60. [2] Hoei-Hansen CE, Carlsen E, Jorgensen N, et al. Towards a non-invasive method for early detection of testicular neoplasia in semen samples by identification of fetal germ cell- specific markers. Hum Reprod 2007;22:167–73. Niels J. van Casteren Gert R. Dohle Leendert H.J. Looijenga*

109 CHAPTER 7 Letter to the Editor

Re: Niels J. van Casteren, Hans Stoop, Gert R. Dohle, Ronald de Wit, J. Wolter Oosterhuis, Leendert H.J. Looijenga. Non-invasive Detection of Testicular Carcinoma In Situ in Semen Using OCT3/4. Eur Urol 2008;54:153–60 We have with interest read the paper by van Casteren et al [1], the subsequent letter by Dieckmann [2], and the author’s reply [3], and consider this line of research commendable. The study is a continuation of many years of research aiming at non-invasive detection of testicular carcinoma in situ (CIS) [4].

The primary objective of van Casteren et al was to evaluate the suitability of OCT- 3/4 for immunocytochemical detection of CIS cells in 56 selected participants. The study is a confirmatory study of our earlier study [5], where we tested another solid marker for CIS, AP-2y, in semen from 503 individuals, and, in parallel, examined OCT-3/4 in 84 participants [5]. Our conclusion was that AP-2y and OCT-3/4 are equally suitable for CIS detection in semen [5]. In patients with a testicular germ cell tumour before orchidectomy we detected OCT-3/4- positive/borderline-positive cells in 35.3%, similar to AP-2y (33.9%). With OCT-3/4 we analysed one patient with CIS-only and did not detect any positive cells. In van Casteren et al [1] three patients with CIS-only were examined and in all cases positive cells were found. If one considers the number of patients (3/3 vs 0/1), there is no scientific or statistical basis to state that the OCT-3/4 assay ‘‘failed’’ in our hands [2]. Nor is there, based on the data presented by Van Casteren et al [1], any scientific basis for their statement that OCT-3/4 is the ‘‘optimal target’’ for CIS detection. Van Casteren et al. state in the paper that their data are preliminary, but still draw extensive conclusions.

We believe that the two studies basically are not comparable, as their designs and inclusion criteria were completely different. The van Casteren et al [1] study was performed on a selected group of patients where CIS or an overt tumour was suspected, whereas we performed a screening of a total of 503 andrologic patients, only a quarter of which had germ cell neoplasia [5]. Van Casteren et al obviously had a much higher likelihood of identifying positive samples, when the study group consisted of 15/56 shown to have CIS (26.8%) compared to our study consisting of 73/503 (14.5%). The larger group examined by us also explains why we detected three false-positive cases, whereas van Casteren et al detected none, as we included 292 controls and they 15. We believe that it cannot from such a small study be concluded that exfoliated CIS cells can be detected in 80–100% of cases [1], especially if no statistical analysis of predictive values or confidence intervals are given.

110 NON-INVASIVE DETECTION OF TESTICULAR CIS IN SEMEN

We agree with Dieckmann [2] that there are several instances where a semen- based analysis is not applicable, that is, for detection of contralateral CIS. However, a semen-based method would be of great use in screening where CIS is suspected, and semen analysis could aid in selection of patients for biopsy, thereby decreasing the number of unnecessary surgeries. We agree here with van Casteren and colleagues that there is a large potential for a non-invasive detection method, best illustrated by the detection of CIS by the AP-2y-based method in a young subfertile man, where testicular neoplasia was detected at the non-invasive stage due to his inclusion as a control subject in our studies [5].

Conflicts of interest: The studies of the Copenhagen group on various aspects of testicular cancer are supported by the Danish Medical Research Council, Vissing Foundation, Danish Cancer Society, Carla Thiel Kraghs Foundation, Kirsten & Freddy Johansen’s Foundation, Danish National Advanced Technology Foundation, and Copenhagen University Hospital. A patent application covering the use of novel CIS markers for diagnosis of germ cell neoplasia is pending.

References [1] van Casteren NJ, Stoop H, Dohle GR, de Wit R, Oosterhuis JW, Looijenga LHJ. Non-invasive detection of testicular carcinoma in situ in semen using OCT3/4. Eur Urol 2008;54:153–60. [2] Dieckmann K-P. Re: Niels J. van Casteren, Hans Stoop, Gert R. Dohle, Ronald de Wit, J. Wolter Oosterhuis, Leen-dert H.J. Looijenga. Non-invasive detection of testicular carcinoma in situ in semen using OCT3/4. Eur Urol 2008;54:153–60. Eur Urol 2009;55:e63–4. [3] van Casteren NJ, Dohle GR, Looijenga LHJ. Reply to Klaus-Peter Dieckmann’s Letter to the Editor re: Niels J. van Casteren, Hans Stoop, Gert R. Dohle, Ronald de Wit, J. Wolter Oosterhuis, Leendert H.J. Looijenga. Non-invasive detection of testicular carcinoma in situ in semen using OCT3/4. Eur Urol 2008;54:153–60. Eur Urol 2009;55:e65–6. [4] Hoei-Hansen CE, Olesen IA, Jorgensen N, et al. Current approaches for detection of carcinoma in situ testis. Int J Androl 2007;30:398–404. [5] Hoei-Hansen CE, Carlsen E, Jorgensen N, Leffers H, Skak-kebæk NE, Rajpert-De Meyts E. Towards a non-invasive method for early detection of testicular neoplasia in semen samples by identification of foetal germ cell-specific markers. Hum Reprod 2007;22:167– 73.

Christina E. Hoei-Hansen* Kristian Almstrup Niels Erik Skakkebæk Ewa Rajpert-De Meyts University Department of Growth and Reproduction (GR-5064), Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark

111

A:IM>Kß

ÔÚL84E B?7JB@4AJ

8EB8A78BA: 4AF&GBBC 4EG

Clinical history A 40-year-old woman presented with blurred vision and diplopia, followed by slowly progressive left-sided motor and sensory disturbances. She also suffered from memory loss and mild spatial and temporal disorientation. A T2-weighted MRI (Figure 1A) showed a large area of high signal intensity (indicated by arrows) in the periventricular white matter of the right more than the left occipital region and the corpus callosum, without enhancement on T1-weighted images (Figure 1B), There was no mass effect. A tumor was considered unlikely, and acute demyelinating encephalomyelopathy was suspected. Repeated cerebrospinal fluid (CSF) examinations showed a mononuclear pleocytosis (10 cells per cubic mm, normal value: <3cells), without immunohistochemical evidence of tumor cells; tumor markers HCG, AFP and CEA were not elevated. Flow cytometry showed predominantly reactive T-lymphocytes, but again no evidence of tumor cells. CT scan of thorax and abdomen did not reveal any abnormalities. The CSF- serum IgG index was normal, and oligoclonal bands were absent. No evidence of a vasculitis or an inflammatory disease was found. Treatment with steroids proved ineffective.

Microscopic description A stereotactic biopsy of the intracerebral lesion showed blast-like neoplastic cells within a mononuclear infiltrate (Figure 2A). The rounded tumor cells contained large centrally located nuclei. A wide panel of markers including those for carcinoma, melanoma and primary central nervous system lymphoma was applied: CD3, CD4, CD5, ALK-1, CD19, CD20, CD79a, CD45, CD30, S-100, MELAN-A, HMB45, CD68, CD43, placental alkaline phosphatase (PLAP), hCG, AFP and CD56, all found to be negative; there was some punctuated NCL5D3 (low molecular weight keratins 8 and 18) positivity (Figure 2B). No final diagnosis could be made.

Figure 1. (A) T2-weighted MRI. Large area of high signal intensity (indicated by arrows). (B) T1-weighted MRI.

114 CASE REPORT: A 40-YEAR OLD WOMEN WITH A PROGRESSIVE WHITE MATTER LESION Subsequently, the marker OCT3/4 became available, which has proven to be specific for certain histological types of germ cell tumors (6), including seminomatous tumors and embryonal carcinoma. This has been confirmed in multiple independent studies (1) (for review). One hundred per cent of tumor cell nuclei present in the biopsy of the above-mentioned patient clearly stained positive for OCT3/4 (Figure 2C). Because only a small number of tumor cells were present in this slide and no biopsy material was left anymore, OCT3/4 was also applied to a slide previously found to be negative by immunohistochemistry for another unrelated marker. Again all tumor nuclei stained positive for OCT3/4 without any background. The stem cell factor receptor c-KIT was also applied to a previously negative slide and the cytoplasm of some tumor cells was stained positive (Figure 2D).

Figure 2. Tumour localized in the brain. (A) H&E staining. Blast-like neoplastic cells within a mononuclear infiltrate. (B) some punctuated NCL5D3 positivity (brown). (C) OCT3/4 positive nuclei (brown). (D) c-KIT positive cytoplasm (red).

115 CHAPTER 8 Diagnosis Diffusely infiltrating germinoma.

Additional treatment and clinical progress After the lesion was diagnosed as a germinoma, the patient was treated with whole brain irradiation in fractions of 1.6 Gy, to a total dose of 44.8 Gy. Follow-up MRI showed disappearance of the enhancement and her condition improved, although focal deficits remained present. Three years after treatment she had a relapse which was treated with chemotherapy (bleomycin-etoposide-cisplatin) with good response. One year later a second relapse occurred.

Discussion Germinomas are tumors of germ cell origin and represent the brain counterpart of testicular seminoma. They are relatively rare and represent 0.4%–3.4% of all primary intracranial tumors, although in Northeast Asia this is significantly higher, up to 9% (4). Like other extragonadal germ cell tumors, germinomas occur primarily in the midline, with the region of the pineal gland being the most common site, followed by the suprasellar compartment (8). They occur predominantly in children with a peak incidence around 12 years of age (4).

Depending on anatomical localization, patients with germinoma may present with a range of neurological, endocrine and psychiatric symptoms. Pineal region tumors often compress the cerebral aqueduct, resulting in progressive hydrocephalus, and invade the tectal plate, producing a paralysis of upward gaze and convergence known as Parinaud syndrome. Suprasellar germ cell tumors compress the optic chiasm, resulting in visual loss, and present with endocrinopathy caused by pituitary failure with diabetes insipidus, retarded growth and sexual maturation (8). Cure rates for this tumor exceed 90% at 10 years, and limitation of treatment-related late morbidity is therefore essential (7).

In this case, the characteristic radiological features of a germinoma were absent (3, 5). There was no mass effect, that is, compression of pre-existent anatomical structures by growth of a solid tumor, and no enhancement after gadolinium administration. Instead, the tumor was diffusely infiltrating the surrounding tissue and followed the contours of the ventricular system. In addition, the marker PLAP usually present in germinomas (2) was negative and diagnosis was made on nuclear OCT3/4 staining, which has proven to be a valuable marker in diagnosis of germ cell tumors (1, 6).

116 CASE REPORT: A 40-YEAR OLD WOMEN WITH A PROGRESSIVE WHITE MATTER LESION References 1. De Jong J, Stoop H, Dohle GR, Bangma CH, Kliffen M, Van Esser JWJ et al (2005) Diagnostic value of oct3/4 for pre-invasive and invasive testicular germ cell tumours. J Pathol 206:242–249. 2. Felix I, Becker LE (1990) Intracranial germ cell tumors in children: an immunohistochemical and electron microscopic study. Pediatr Neurosurg 16:156–162. 3. Fujimaki T, Matsutani M, Funada N, Kirino T, Takakura K, Nakamura O et al (1994) CT and MRI features of intracranial germ cell tumors. J Neurooncol 19:217–226. 4. Jennings MT, Gelman R, Hochberg F (1985) Intracranial germ-cell tumors: natural history and pathogenesis. J Neurosurg 63:155– 167. 5. Kanagaki M, Miki Y, Takahashi JA, Shibamoto Y, Takahashi T, Ueba T et al (2004) MRI and CT findings of neurohypophyseal germinoma. Eur J Radiol 49:204–211. 6. Looijenga LHJ, Stoop H, De Leeuw HPJC, De Gouveia Brazao CA, Gillis AJM, Van Roozendaal KEP et al (2003) POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 63:2244–2250. 7. Rogers SJ, Mosleh-Shirazi MA, Saran FH (2005) Radiotherapy of localised intracranial germinoma: time to sever historical ties? Lancet Oncol 6:509–519. 8. Rosenblum MK, Matsutani M, Van Meir EG (2000) CNS germ cell tumours. In: World Health Organization Classification of Tumours. Pathology and Genetics of the Tumours of the Nervous System. P Kleihues, WK Cavenee (eds), pp. 208–214. IARC Press, Lyon

117

A:IM>Kà

<998E8AG<4?8KCE8FF

BHEA4?B9#4G;B?B:LÒÛß ÒÛ ÓÚ ÒÚÚÙ

8EB8A78BA: 4AF&GBBC 7

Abstract Combined action of SOX and POU families of transcription factors plays major roles in embryonic development. In embryonic stem cells, the combination of SOX2 and POU5F1 (OCT3/4) is essential for maintaining the undifferentiated state by activating pluripotency-linked genes, and inhibition of genes involved in differentiation. Besides embryonic stem cells, POU5F1 is also present in early germ cells, primordial germ cells, and gonocytes, where it has a role in suppression of apoptosis. Here we demonstrate that SOX2 is absent in germ cells of human fetal gonads, and as expected carcinoma in situ (CIS), ie the precursor lesion of testicular germ cell tumours of adolescents and adults (TGCTs), and seminoma. Based on genome-wide expression profiling, SOX17 was found to be present, instead of SOX2, in early germ cells and their malignant counterparts, CIS and seminoma. Immunohistochemistry, western blot analysis, and quantitative RT-PCR showed that SOX17 is a suitable marker to distinguish seminoma from embryonal carcinoma, confirmed in representative cell lines. Aberrant SOX2 expression can be present in Sertoli cells when associated with CIS, which can be misdiagnosed as embryonal carcinoma. In conclusion, this study demonstrates the absence of SOX2 in human embryonic and malignant germ cells, which express SOX17 in conjunction with POU5F1. This finding has both diagnostic and developmental biological implications. It allows the identification of seminoma-like cells from embryonal carcinoma based on a positive marker and might be the explanation for the different function of POU5F1 in normal and malignant germ cells versus embryonic stem cells.

120 SOX17 AND SOX2 IN GERM AND STEM CELLS Introduction The SOX family of transcription factors is involved in orchestrating development from the early cells in the embryo to differentiated lineages of specialized cells. Currently, 20 SOX proteins have been identified in humans [1]. SOX proteins have similarity to the HMG domain of SRY, the testis-determining factor in mammals excluding monotremes [2], and are therefore named SRY-related HMG box (SOX) proteins. Although different SOX proteins can bind highly similar DNA sequences in vitro, they are expressed in a cell-type-specific manner by which they regulate their target genes [3]. This regulation is achieved by the collaboration of SOX proteins in association with other proteins [4]. Abundantly represented among the protein partners are the POU homeodomain proteins [5,6]. These transcription factors are also expressed in a cell-specific way and interact simultaneously with the DNA and stabilizing partner proteins [7,8]. The cooperative activity of SOX and POU proteins has been confirmed in different systems, such as lens placode formation [9], myelination of Schwann cells [10], and in oligodendrocytes [11].

The most extensively studied SOX–POU partnership is that of SOX2 and POU5F1 (also known as OCT3/4 and OCT4) in the inner cell mass of early embryonic development and the in vitro counterpart, ie embryonic stem (ES) cells [12,13– 15]. Both factors are in the centre of an intricate network required to maintain pluripotency (see refs 16–18 for a review), where feedback mechanisms control the appropriate level of the transcription factors [19], and deviation will promote differentiation [20,21]. Genome-wide screens in human and mouse ES cells have identified a large number of target genes involved in the regulation of pluripotency and differentiation [22,23]. Besides being present in ES cells, POU5F1 is also present in primordial germ cells (PGCs), the cell of origin of gametogenesis, which arise from the proximal epiblast cells of the bi-layered embryo [24,25]. During the migration of PGCs, POU5F1 remains highly expressed [26] and after arrival at the gonadal ridge (the PGCs are now called gonocytes), these cells progressively lose POU5F1 upon maturation during gametogenesis. Interestingly, knock-out of POU5F1 in mouse PGCs results in induction of apoptosis, instead of differentiation as in ES cells [27]. It is not known so far how this functional difference between POU5F1 in ES and PGC cells, which might be related to a different interaction with a SOX member (see below), is regulated. Normally, 1 year after birth, no POU5F1-positive cells are present either in the gonads or in any other organs and tissues. In humans, prolonged expression of POU5F1 is strongly associated with the development of malignant testicular germ cell tumours of adolescents and young adults (TGCTs), which clinically manifest after puberty. These so-called type II TGCTs, ie seminomas and non- seminomas, arise from a malignantly transformed PGC/gonocyte and this

121 CHAPTER 9 dormant precursor lesion is named carcinoma in situ (CIS) or intratubular germ cell neoplasia unclassified (IGCNU) (see ref 28 for a review). The precursor lesion of dysgenetic gonads is known as gonadoblastoma (GB) [29,30]. After puberty, proliferation of these precursor cells will give rise to invasive seminoma (known as dysgerminoma in the ovary and dysgenetic gonad), the default pathway for CIS cells, or - by acquiring additional expression of ES cell markers - to invasive embryonal carcinoma (EC), the undifferentiated cells of non-seminoma, representing the inner cell mass during normal embryogenesis. Furthermore, EC cells can differentiate into teratoma (the somatic lineage) and the different extra-embryonal lineages, ie yolk sac tumour and choriocarcinoma, as well as the germ cell lineage [31]. POU5F1 appears to be a specific marker in TGCT diagnostics with 100% of CIS, seminoma, and EC cells, showing profound nuclear staining [32–36].

The role of SOX2 interaction with POU5F1 in maintaining pluripotency and preventing differentiation is well established [13]; however, how POU5F1 expression is involved in apoptosis control in PGCs (see above) is not clear. This cannot be due to NANOG, for example, because this protein is expressed in PGCs/gonocytes and ES, as well as in their malignant counterparts CIS/GB, seminomas, and EC [37]. The presence of both NANOG and POU5F1, and the absence of SOX2 have been proposed to be informative for distinguishing seminoma-like cells from EC. However, a specific positive marker for the seminoma-like cells would be of interest in a diagnostic setting, like we previously reported for spermatocytic seminoma compared with (classic) seminoma [38]. We show here that SOX2 is not present in germ cells of early embryonic gonads to adult testis and ovary, or in pre-invasive and invasive TGCTs by immunohistochemistry, but we have confirmed the presence in EC, as well as heterogeneously in different derivatives. SOX2 can also be present in Sertoli cells associated with CIS; this observation is novel and significant, because it may result in overdiagnosis of intratubular EC, as done recently [39]. Moreover, this study shows that SOX17, instead of SOX2, is present in CIS and seminoma cells. Although a differential pattern of expression of SOX17 has been reported before [40,41], this study for the first time investigates SOX17 in more detail, especially on the protein level, in the context of both normal germ cell development and germ cell tumour diagnosis. SOX17 has multiple functions [42–44], including a specific role in embryonic compared with adult haematopoietic stem cells [45]. Immunohistochemistry of SOX17 demonstrated nuclear co-expression with POU5F1 in fetal gonocytes, CIS, and seminoma cells, as well as in the recently characterized seminoma cell line TCam-2, but not in EC and derived cell lines. This finding allows distinction between seminoma and EC in a diagnostic setting, for which so far no absolute positive marker for seminoma is available. In addition,

122 SOX17 AND SOX2 IN GERM AND STEM CELLS this knowledge will improve our understanding of the progression of TGCTs from CIS to either seminoma or EC.

Materials and methods Research on human gonadal and tumour samples has been performed according to the Code for Proper Secondary Use of Human Tissue in The Netherlands, as developed by the Dutch Federation of Medical Scientific Societies (FMWV) (version 2002), and has been approved by an institutional review board (MEC 02.981). All samples were collected in close collaboration with (paediatric) urologists and pathologists in the southwestern part of The Netherlands. The following normal gonadal samples were used: 16 samples of fetal testis (15–40 weeks’ gestational age), five fetal ovaries (18–41 weeks’ gestational age), 42 adult testes, and four adult ovaries. Type II TGCTs were as follows: 39 CIS samples, 50 seminomas, 42 ECs, eight yolk sac tumours, 13 choriocarcinomas, and 26 teratomas. Nine dysgenetic gonads - some containing gonadoblastoma or dysgerminoma - were used for SOX2 immunohistochemistry (see below).

Immunohistochemistry The presence of SOX2 in normal and pathological gonadal development was studied by immunohistochemistry using a monoclonal antibody (AF2018; R&D Systems, Abingdon, UK) in a 1 : 750 dilution incubated for 2 h at room temperature. As a secondary antibody, a biotinylated rabbit anti-mouse antibody was applied (1 : 200; E0413; Dako, Glostrup, Denmark) and bound antibody complex was visualized using the horseradish peroxidase avidin–biotin complex (HRP-ABC) method. The samples examined are listed above. In addition, double staining was performed using a monoclonal antibody directed against the N- terminus of the human POU5F1 (sc-5279; 1 : 1000; sc-Santa Cruz Biotechnology, Santa Cruz, CA, USA) and SOX2, respectively, with 3-amino-9-ethyl-carbazole

(A5754; Sigma, Steinheim, Germany)/H2O2 for red staining and Fast Blue/naphthol AS-MX phosphate (F3378 and N500; Sigma) for blue staining. Detection of SOX17 in gonadal development, normal spermatogenesis, and TGCTs was carried out with a goat polyclonal antibody (GT15094; Neuromics, Edina, MN, USA) in a 1 : 3000 dilution for 2 h at room temperature. Subsequently, a biotinylated horse anti-goat secondary antibody (1 : 200; BA-9500; Vector) was used and detection was carried out using the HRP-ABC method. Antibodies against SOX17 from SantaCruz (sc-17 355) and R&D Systems (MAB1924) were also tested, but these gave non-specific and cytoplasmic staining, tested by western blotting and immunohistochemistry, respectively. These antibodies were therefore not used for further experiments.

123 CHAPTER 9 In addition to the fetal gonads indicated above, a series of early fetal gonads (57– 63 days’ gestational age) was obtained following medical termination of pregnancy [46] and fixed in Bouin’s. Extensive testing using an appropriate positive control showed that the SOX2 antibody (AF-2018; R&D Systems) did not work on Bouin-fixed material and therefore only SOX17 (GT15094; Neuromics) immunohistochemistry could be performed on these samples.

Cell culture The cell line TCam-2, recently confirmed to be a seminoma cell line [47], was obtained from Sohei Kitazawa (Division of Molecular Pathology, Kobe University, Japan). The cells were cultured in RPMI 1640 (Invitrogen, Breda, The Netherlands) supplemented with 10% fetal calf serum (Hyclone, Thermo Fisher Scientific, Etten-Leur, The Netherlands) and penicillin/streptomycin (Invitrogen, Breda, The Netherlands) at 370C in a humidified cell culture incubator with 5% carbon dioxide. The EC cell lines Tera-1 [48], NTera-2 [49], 2102Ep [50], 833KE [51], and NCCIT [52] were cultured as described before [53]. The cell line JKT-1, recently demonstrated not to be a seminoma cell line, was cultured as described before [54]. The cells likely represent a later stage of germ cell development. Cells were trypsinized and harvested for the different experiments. Cytospins were prepared and RNA was isolated (see below).

Protein isolation and western blot analysis Isolation of protein and western blot analysis were essentially performed as described previously [33]. The antibodies used were the same as those used for immunohistochemistry. In addition, mouse monoclonal beta-actin (clone AC-15; Sigma Aldrich, St Louis, MO, USA) was used. Binding of the primary antibodies was visualized by using IRDye donkey anti-mouse or donkey anti-goat secondary antibodies and the blots were scanned on the Odyssey infrared imaging system (all from LI COR Biosciences, NE, USA).

Quantitative RT-PCR for SOX2 and SOX17 in normal testis and TGCTs High-quality total RNA was extracted from normal testis (n = 4) and patient- derived TGCT samples con-taining either CIS (n = 4) or seminoma (n = 12) or EC (n = 11) using an RNAqueous-4PCR kit (Ambion Europe, Huntingdon, UK) according to the manufacturer’s instructions. In addition, RNA from the cell lines JKT-1, TCam-2, and the EC cell lines Tera-1, NTera-2, 2102Ep, and NCCIT was used. Samples were treated with DNase and checked for residual DNA contamination by PCR. Quantitative PCR was performed using the real-time PCR ABI PRISM 7700 sequence detector system (Applied Biosystems, Foster City, CA, USA). For the detection of SOX2 expression, the following primer set was used: SOX2-F 5ǯ - ATGGAGACGGACCTGAAGC-3’ and SOX2-R 5’-GCAGCGTGTACTTATCCTTC-3’. Quantitative values were obtained from the Ct. SOX2 mRNA was quantified

124 SOX17 AND SOX2 IN GERM AND STEM CELLS relative to HPRT [SOX2 mRNA value = 2(mean Ct HPRT−mean Ct SOX2)] [38]. For the detection of SOX17, the following primers were used: SOX17-F (exon 1) 5’-GATGCGGGAT ACGCCAGTGAC-3’ and SOX17-R (exon 2) 5’-GCTCTGCCTCCTCCACGAAG-3’ and SOX17 mRNA was quantified by the formula described above.

Micro-array expression profiling The expression profiles of frozen TGCTs (five spermatocytic seminomas, three dysgerminomas, being the ovarian counterpart of seminoma [38], four seminomas, and five ECs), the cell line JKT-1, and the seminomatous TCam-2 cell line were determined using the Affymetrix GeneChip Human Genome U133 plus 2.0 array as described before [38]. Analysis was done using Omniviz and SAM software as reported in the same paper. Expression levels of each SOX family member were analysed for all samples and plotted as a graph.

Results

SOX2 is not expressed in normal human gonadogenesis To determine whether SOX2 may play a role in normal germ cell development and gonadogenesis, immunohistochemistry was performed on gonads from early fetal life (age: 15 weeks’ gestational age and older) to adult ovary and testis. The SOX2 protein was not expressed in germ cells, regardless of their state of maturation (data not shown). Occasionally, SOX2-positive Sertoli cells were present in normal adult testis; however, the expression level of SOX2 mRNA was very low (Figure 1A).

SOX2 in testicular germ cell tumours The differential expression of SOX2 between seminoma and EC cells is used as an argument to explain the differences in pluripotency between these cells [41,55]. In contrast to SOX2, the other markers related to pluripotency, ie POU5F1 and NANOG, show a similar pattern of expression in these histological subtypes. The relative overexpression of several core ‘stemness’ genes, such as EBAF, TDGF1, and SOX2, reflects the undifferentiated state of EC cells and is the starting point of their differentiation potential, analogous to the ES cells derived from the inner cell mass of early embryonic development.

To study SOX2 expression in the development of TGCTs, tissue micro-arrays (TMAs) of the various invasive histological subtypes and the precursor lesion CIS were used. As expected, also based on previous studies [56], all EC cells were positive for SOX2, while CIS, seminoma, yolk sac tumour, and choriocarcinoma were consistently negative. In teratoma, a small number of SOX2-positive cuboidal or columnar epithelium cells of endodermal origin were present, as was also recently reported in teratomas of the central nervous system [57]. However,

125 CHAPTER 9 in the seminiferous tubules containing CIS cells (always POU5F1-positive), a large number of SOX2-positive cells were seen to be more luminally situated than the CIS cells (Figure 1B). This observation has not been reported before and deserves further investigation. Detailed morphological study of these SOX2-positive cells and double staining for SOX2 and vimentin showed that they are in fact Sertoli cells, and not tumour cells (Figure 1C). Indeed, quantitative RT-PCR confirmed the elevated expression of SOX2 in samples containing CIS cells, compared with normal testis (Figure 1A). This presence of SOX2 in the context of germ cell tumours might be overdiagnosed as EC, when per-formed as single staining. However, when it is combined with staining for POU5F1, it must be investigated whether SOX2 and POU5f1 are present within the same cells, indicating EC, preferentially using a double-staining approach.

Figure 1. (A) Quantitative RT-PCR for SOX2 on TGCT samples. The cell line JKT-1, the seminomatous cell line TCam-2, and the EC cell lines (NTera-2, NCCIT, Tera-1, and 2102Ep) were tested. High expression of SOX2 was observed in the EC samples and cell lines. In testis parenchyma containing CIS (the percentage is indicated in the bars), SOX2 mRNA expression was elevated compared with normal testis; the level of expression parallels the amount of CIS-containing seminiferous tubules. TCam-2 and JKT-1 show a level of expression in between normal testis/seminoma and EC samples and cell lines. (B) Double immunohistochemistry for SOX2 (red) and POU5F1 (blue). SOX2 is not expressed in the POU5F1-positive CIS cells, but in a cell type more luminally located, being Sertoli cells. (C) Double staining for vimentin (blue) and SOX2 (red) combined with morphology and the position in the seminiferous tubule shows that the SOX2-positive cells are Sertoli cells

126 SOX17 AND SOX2 IN GERM AND STEM CELLS SOX17 is expressed in gonocytes, CIS/gonadoblastoma, and seminoma/dysgerminoma cells With the absence of SOX2 expression in germ cells and their malignant counterparts, CIS and seminoma cells, the question arises as to whether another SOX member will be specifically expressed in these cells. This might be related to the different regulatory function of POU5F1 in the germ cells versus the stem cells, ie apoptosis and differentiation. To test this, we screened the expression profiles of all SOX family members in TGCT samples and the cell lines JKT-1 and TCam-2 (Figure 2). The expression pattern of SOX2 was confirmed on the mRNA level (being specifically expressed in EC). Most of the SOX members gave a non- specific pattern for the different histological elements investigated, including the cell lines. Possible exceptions are SOX15 and SOX17. For this study, we focused on SOX17, although SOX15 is of interest as well. SOX17 was highly expressed in all seminoma/dysgerminoma samples, as well as the seminoma cell line TCam-2, and virtually absent in EC cells (absolute value for specific probe sets below 30). The mRNA expression profile of SOX17 in POU5F1-positive TGCTs (seminoma /dysgerminoma and EC) was exactly complementary to that of SOX2. Quantitative RT-PCR confirmed the elevated SOX17 expression in seminoma samples and the seminoma cell line TCam-2, compared with EC samples and cell lines (Figure 3A). Immunohistochemistry using a specific antibody showed the nuclear expression of SOX17 in all CIS/gonadoblastoma and seminoma/ dysgerminoma cells, being co-expressed with POU5F1 (Figures 3B and 3C). In addition, the seminoma cell line TCam-2 expressed SOX17 protein (Figures 3D and 3E), while no SOX17 was detected on cytospins of EC cell lines (data not shown) or by western blot analysis (Figure 3E). No staining was identified in either Sertoli cells (testis) or granulosa cells (ovary).

The specific nuclear staining for SOX17 is of diagnostic value in discriminating seminoma/dysgerminoma from EC. The differential mRNA expression of SOX17 between seminoma/dysgerminoma and spermatocytic seminoma, as shown in Figure 2, is in line with earlier findings [38]. In spite of its value in distinguishing seminoma from embryonal carcinoma based on a positive marker in the first, SOX17 is not a suitable marker for the diagnosis of CIS, because it is found in different maturation stages of spermatogenesis (Figure 3F). In addition, SOX17 expression is present in fetal male and female gonocytes (the earliest sample tested was a 57 days old gestational age gonad) and remains present in these cells throughout fetal life (Figure 3G). In contrast to male spermatogenesis, SOX17 is not detected in oogenesis beyond the stage of the gonocyte. However, SOX17 protein is detected in the epithelium of the Fallopian tube and rete testis (data not shown). Therefore POU5F1 is still the best marker for the detection of CIS in the adult testis [33]. However, SOX17 can be helpful in distinguishing

127 CHAPTER 9 between the invasive components seminoma and EC, especially when performed together with analysis of POU5F1.

Discussion Transcription factors act as pathway switches in embryonic development and commitment to the differentiation lineages. Specific combinations of transcription factors belonging to the SOX and POU families are known to establish tissue and cell-specific expression of target genes. In this study, we focused on possible protein partners of POU5F1 which are subsequently expressed during embryonic development in cells of the inner cell mass, the epiblast and PGCs. In ES cells, the intimate relationship between POU5F1 and SOX2 has been most extensively studied and the binding of both proteins to adjacent sites on the DNA with simultaneous physical protein interaction controls a large number of target genes involved in pluripotency and differentiation. However, in PGCs and gonocytes, SOX2 is lacking (this paper), and POU5F1 has a regulatory role in apoptosis, for which the mechanism is unknown so far.

Besides ES cells, SOX2 is abundantly expressed in differentiating and differentiated cells and is found in endodermal cells of the primitive foregut [58], the developing inner ear [59], and the central nervous system [60]. However, in these cells, it is never detected in combination with POU5F1, which is likely related to the absence of pluripotency. It has been reported that SOX2 is expressed in mouse fetal germ cells and neonatal ovine gonads [61,62], but the actual data were not shown. These reports were confirmed by RNA in situ detecting SOX2 in male E12.5 to E15.5 gonad/mesonephros tissues; however, no protein detection was done [63]. Recently, the expression of SOX2 in the establishment and early migration of mouse PGCs was identified by quantitative single-cell gene expression profiling [64]. These results indicated that SOX2 expression is not immediately lost upon commitment to the germ cell lineage. However, the timing of SOX2 down-regulation in gametogenesis remains to be clarified. Our results show that SOX2 is not expressed in human gonadogenesis from 15 weeks of gestation onwards, or in normal adult testis and ovary.

The study in early mouse PGCs also investigated the expression pattern of SOX3 and SOX17 [64]. Both genes were transiently expressed in migrating PGCs at developmental stage E7.25 dpc. Here we demonstrated for the first time that SOX17 is expressed and the protein is present in CIS and seminoma cells, but not in EC cells, either in vivo tumours or cell lines. This is in line with previous studies, which also demonstrated SOX17 as differentially expressed between seminoma and embryonal carcinoma at the mRNA level, although no follow-up studies were performed [40,41]. The absence of SOX3 expression in seminoma cells (Figure 2)

128 SOX17 AND SOX2 IN GERM AND STEM CELLS could be related to the timing of the maturation block that has occurred in the precursor cells of TGCTs, or be the result of species-specific differences. So far, the existing evidence for a role of SOX17 in gametogenesis is fragmented. Heterozygous SOX17 knock-out mice are normal and fertile, while homozygous knock-out mice are deficient in gut endoderm and after 10.5 dpc, no homozygous mutants are found [65]. This lethality due to endoderm deficiency is prior to gonad development and therefore the role of SOX17 absence during gametogenesis can be studied only in conditional knock-outs. Furthermore, two isoforms (one without an HMG domain) have been isolated from an adult mouse testis library. Analysis of these isoforms suggests that SOX17 may function as a transcriptional activator in pre-meiotic germ cells, and that a splicing switch into a truncated SOX17 protein may lead to loss of its function in post-meiotic germ cells [43]. So far, no human SOX17 splice variants have been identified. The co- expression of SOX17 and POU5F1 in CIS and seminoma cells in the absence of SOX2 qualifies SOX17 as a potential protein partner in these cells, which is currently under investigation. This observation initiates a new hypothesis in the regulation of the function of POU5F1 (see below). In addition, SOX17 is a repressed target gene of the POU5F1 protein and not of SOX2 in human ES cells [22]. It remains to be investigated whether this is also the case in PGCs.

The co-expression of SOX17 and POU5F1 in the seminoma cell line TCam-2 compared with established EC cell lines, being SOX2-positive and SOX17- negative, is an informative experimental model for studying this hypothesis, and for identifying common target genes in CIS and seminoma. One of the interesting hypotheses to test is whether SOX17 is the determinant factor resulting in the regulatory role of POU5F1 in apoptosis in normal and malignant germ cells, while the co-presence with SOX2 regulates differentiation in normal and malignant embryonic stem cells. The WNT pathway is possibly a regulatory factor involved in the transcriptional activity of POU5F1 and SOX17 in normal and malignant germ cells. This has been shown in ES cells [66] and embryonic endoderm [67] in which the proteins physically interact with ș-catenin. A recent observation is that SOX17 is essential for fetal and neonatal haematopoietic stem cells (HSCs), but not for adult HSCs [45]. Knock-out studies and expression data of SOX17 in primitive endoderm, fetal HSCs, and early germ cells indicate a crucial role for this transcription factor in a number of lineages during embryonic and fetal development. The origin of PGCs from the primitive endoderm, like fetal HSCs, is of high interest in this context.

129 CHAPTER 9

Figure 2. Expression results of all 20 human SOX family members in TGCT samples and cell lines JKT-1 and TCam-2 based on genome-wide (Affymetrix) expression profiling (log2 transformed data). The following samples were used: five spermatocytic seminomas, four seminomas, three dysgerminomas, and five ECs (also indicated in the legend in the bottom right corner of the figure explaining the different bar colours). The main findings were the well-known high SOX2 expression in EC and low expression in seminoma/dysgerminoma. The profile of SOX17 shows an inverse relationship with low expression in EC and high expression in seminoma/dysgerminoma samples and the seminoma cell line TCam-2 (indicated by an asterisk)

130 SOX17 AND SOX2 IN GERM AND STEM CELLS

Figure 3. (A) Quantitative RT-PCR for SOX17 on TGCT samples, the cell line JKT-1, the seminomatous cell line TCam-2, and EC cell lines (NTera-2, NCCIT, Tera-1, and 2102Ep) demonstrates high expression of SOX17 in the seminoma samples and cell line TCam-2, whereas the expression is low in all the other samples. Nuclear SOX17 signal (brown) in (B) CIS, (C) seminoma, and (D) the seminomatous cell line TCam-2. Inset: nuclear POU5F1 staining in TCam-2 cells. (E) Western blotting shows a complementary expression pattern for SOX2 and SOX17 in the TCam-2 and EC cell lines with high SOX17 expression in TCam-2 and no SOX17 expression in the EC cell lines NTera-2, NCCIT, and 833KE. SOX2 is highly expressed in the EC cell lines but not in TCam-2. The embryonic stem cell marker OCT3/4 is expressed in all samples. ß -actin was used as a loading control. (F) SOX17 protein is present in various maturation stages of spermatogenesis: spermatogonium (black arrow), secondary spermatocyte (white arrow), and spermatid (black arrowhead), while primary spermatocytes, spermatozoa, and Sertoli cells are negative. (G) Nuclear SOX17 in gonocytes of an 18-week-old fetus

131 CHAPTER 9 As suggested, the presence of expression of SOX17 in CIS could indicate the timing of the first step in the development of TGCTs. Whether these cells are already malignantly transformed at the time of this maturation arrest or acquire malignant potential later in development as a result of an unfavourable micro- environment for immature germ cells is not yet clear. Therefore analysis of these markers in patients with disorders of sex development (DSD) will be informative, because some of them show a significant increased risk for type II GCTs [30]. Our study of SOX2 expression in normal gonadogenesis and TGCTs demonstrated that SOX2 can be present in Sertoli cells in CIS-containing seminiferous tubules, whereas this was rarely observed in tubules without CIS. SOX2 is not expressed in fetal gonads from 15 weeks of gestation onwards, indicating either aberrant expression of SOX2 in Sertoli cells in CIS-positive tubules or possibly persistent expression assuming that SOX2 is expressed in Sertoli cells before 15 weeks. Either way, SOX2 expression in CIS-positive tubules reflects an abnormal micro- environment which might have contributed to the aetiology of TGCT initiation or has a later origin under the influence of the aberrant germ cells. An important implication of this result is that SOX2 expression in seminiferous tubules should not be mistaken for intratubular EC known to be SOX2-positive. This misinterpretation has indeed been made recently, leading to overdiagnosis, and possibly overtreatment [39].

Overall, this study shows the significant value of combined investigation of normal germ cell development in shedding light on the pathogenesis of type II GCTs and vice versa. It adds SOX17 as a positive marker for seminoma-like cells to the spectrum of diagnostic markers, which will initiate a new line of research on the regulation of apoptosis and differentiation in the malignant counterparts of PGC and ES, which is so characteristic of this type of cancer.

References 1. Grosschedl R, Giese K, Pagel J. HMG domain proteins: architec-tural elements in the assembly of nucleoprotein structures. Trends Genet 1994;10:94–100. 2. Marshall Graves JA. Sex chromosomes and sex determination in weird mammals. Cytogenet Genome Res 2002;96:161–168. 3. Pevny LH, Lovell-Badge R. Sox genes find their feet. Curr Opin Genet Dev 1997;7:338–344. 4. Wegner M. From head to toes: the multiple facets of Sox proteins. Nucleic Acids Res 1999;27:1409–1420. 5. Dailey L, Basilico C. Coevolution of HMG domains and homeodomains and the generation of transcriptional regulation by Sox/POU complexes. J Cell Physiol 2001;186:315–328. 6. Kamachi Y, Uchikawa M, Kondoh H. Pairing SOX off: with partners in the regulation of embryonic development. Trends Genet 2000;16:182–187. 7. Veenstra GJ, van der Vliet PC, Destree OH. POU domain transcription factors in embryonic development. Mol Biol Rep 1997;24:139–155. 8. Phillips K, Luisi B. The virtuoso of versatility: POU proteins that flex to fit. J Mol Biol 2000;302:1023–1039.

132 SOX17 AND SOX2 IN GERM AND STEM CELLS

9. Donner AL, Episkopou V, Maas RL. Sox2 and Pou2f1 interact to control lens and olfactory placode development. Dev Biol 2007;303:784–799. 10. Ghislain J, Charnay P. Control of myelination in Schwann cells: a Krox20 cis-regulatory element integrates Oct6, Brn2 and Sox10 activities. EMBO Rep 2006;7:52–58. 11. Kuhlbrodt K, Herbarth B, Sock E, Enderich J, Hermans-Borgmeyer I, Wegner M. Cooperative function of POU pro-teins and SOX proteins in glial cells. J Biol Chem 1998; 273:16050– 16057. 12. Remenyi A, Lins K, Nissen LJ, Reinbold R, Scholer HR, Wilmanns M. Crystal structure of a POU/HMG/DNA ternary complex suggests differential assembly of Oct4 and Sox2 on two enhancers. Genes Dev 2003;17:2048–2059. 13. Chew JL, Loh YH, Zhang W, Chen X, Tam WL, Yeap LS, et al. Reciprocal transcriptional regulation of Pou5f1 and Sox2 via the Oct4/Sox2 complex in embryonic stem cells. Mol Cell Biol 2005;25:6031–6046. 14. Okumura-Nakanishi S, Saito M, Niwa H, Ishikawa F. Oct-3/4 and Sox2 regulate Oct-3/4 gene in embryonic stem cells. J Biol Chem 2005;280:5307–5317. 15. Catena R, Tiveron C, Ronchi A, Porta S, Ferri A, Tatangelo L, et al. Conserved POU binding DNA sites in the Sox2 upstream enhancer regulate gene expression in embryonic and neural stem cells. J Biol Chem 2004;279:41846–41857. 16. Wang J, Rao S, Chu J, Shen X, Levasseur DN, Theunissen TW, et al. A protein interaction network for pluripotency of embryonic stem cells. Nature 2006;444:364–368. 17. Rao S, Orkin SH. Unraveling the transcriptional network controlling ES cell pluripotency. Genome Biol 2006;7:230. 18. de Jong J, Looijenga LH. Stem cell marker OCT3/4 in tumor biology and germ cell tumor diagnostics: history and future. Crit Rev Oncog 2006;12:171–203. 19. Pan G, Li J, Zhou Y, Zheng H, Pei D. A negative feedback loop of transcription factors that controls stem cell pluripotency and self-renewal. FASEB J 2006;20:1730–1732. 20. Niwa H, Miyazaki J, Smith AG. Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nature Genet 2000;24:372–376. 21. Masui S, Nakatake Y, Toyooka Y, Shimosato D, Yagi R, Taka-hashi K, et al. Pluripotency governed by Sox2 via regulation of Oct3/4 expression in mouse embryonic stem cells. Nature Cell Biol 2007;9:625–635. 22. Boyer LA, Lee TI, Cole MF, Johnstone SE, Levine SS, Zucker JP, et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 2005;122:947–956. 23. Loh YH, Wu Q, Chew JL, Vega VB, Zhang W, Chen X, et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nature Genet 2006;38:431–440. 24. Avilion AA, Nicolis SK, Pevny LH, Perez L, Vivian N, Lovell-Badge R. Multipotent cell lineages in early mouse development depend on SOX2 function. Genes Dev 2003;17:126–140. 25. Rosner MH, Vigano MA, Ozato K, Timmons PM, Poirier F, Rigby PW, et al. A POU-domain transcription factor in early stem cells and germ cells of the mammalian embryo. Nature 1990;345:686–692. 26. Molyneaux K, Wylie C. Primordial germ cell migration. Int J Dev Biol 2004;48:537–544. 27. Nichols J, Zevnik B, Anastassiadis K, Niwa H, Klewe-Nebenius D, Chambers I, et al. Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4. Cell 1998;95:379–391. 28. Oosterhuis JW, Looijenga LH. Testicular germ-cell tumours in a broader perspective. Nature Rev Cancer 2005;5:210–222. 29. Kersemaekers AM, Honecker F, Stoop H, Cools M, Molier M, Wolffenbuttel K, et al. Identification of germ cells at risk for neo-plastic transformation in gonadoblastoma: an immunohistochemi-cal study for OCT3/4 and TSPY. Hum Pathol 2005;36:512–521. 30. Cools M, Drop SL, Wolffenbuttel KP, Oosterhuis JW, Looi-jenga LH. Germ cell tumors in the intersex gonad: old paths, new directions, moving frontiers. Endocr Rev 2006;27:468–484.

133 CHAPTER 9

31. Honecker F, Stoop H, Mayer F, Bokemeyer C, Castrillon DH, Lau YF, et al. Germ cell lineage differentiation in non-seminomatous germ cell tumours. J Pathol 2006;208:395–400. 32. Looijenga LH, Stoop H, de Leeuw HP, de Gouveia Brazao CA, Gillis AJ, van Roozendaal KE, et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003;63:2244–2250. 33. de Jong J, Stoop H, Dohle GR, Bangma CH, Kliffen M, van Esser JWJ, et al. Diagnostic value of OCT3/4 for pre-invasive and invasive testicular germ cell tumours. J Pathol 2005;206:242– 249. 34. Cheng L, Sung MT, Cossu-Rocca P, Jones T, Maclennan G, De Jong J, et al. OCT4: biological functions and clinical applications as a marker of germ cell neoplasia. J Pathol 2006;211:1– 9. 35. Cheng L, Thomas A, Roth LM, Zheng W, Michael H, Karim FW. OCT4: a novel biomarker for dysgerminoma of the ovary. Am J Surg Pathol 2004;28:1341–1346. 36. Jones TD, Ulbright TM, Eble JN, Baldridge LA, Cheng L. OCT4 staining in testicular tumors: a sensitive and specific marker for seminoma and embryonal carcinoma. Am J Surg Pathol 2004;28:935–940. 37. Hart AH, Hartley L, Parker K, Ibrahim M, Looijenga LH, Pauchnik M, et al.The pluripotency homeobox gene NANOG is expressed in human germ cell tumors. Cancer 2005;104:2092– 2098. 38. Looijenga LH, Hersmus R, Gillis AJ, Pfundt R, Stoop HJ, van Gurp RJ, et al. Genomic and expression profiling of human spermatocytic seminomas: primary spermatocyte as tumorigenic precursor and DMRT1 as candidate chromosome 9 gene. Cancer Res 2006;66:290–302. 39. Berney DM, Lee A, Shamash J, Oliver RT. The association between intratubular seminoma and invasive germ cell tumors. Hum Pathol 2006;37:458–461. 40. Skotheim RI, Lind GE, Monni O, Nesland JM, Abeler VM, Fossa SD, et al. Differentiation of human embryonal carcinomas in vitro and in vivo reveals expression profiles relevant to normal development. Cancer Res 2005;65:5588–5598. 41. Korkola JE, Houldsworth J, Chadalavada RS, Olshen AB, Dobrzynski D, Reuter VE, et al. Down-regulation of stem cell genes, including those in a 200-kb gene cluster at 12p13.31, is associated with in vivo differentiation of human male germ cell tumors. Cancer Res 2006;66:820–827. 42. Sohn J, Natale J, Chew LJ, Belachew S, Cheng Y, Aguirre A, et al. Identification of Sox17 as a transcription factor that regulates oligodendrocyte development. J Neurosci 2006;26:9722– 9735. 43. Kanai Y, Kanai-Azuma M, Noce T, Saido TC, Shiroishi T, Hayashi Y, et al. Identification of two Sox17 messenger RNA isoforms, with and without the high mobility group box region, and their differential expression in mouse spermatogenesis. J Cell Biol 1996;133:667–681. 44. Wang R, Cheng H, Xia L, Guo Y, Huang X, Zhou R. Molecular cloning and expression of Sox17 in gonads during sex reversal in the rice field eel, a teleost fish with a characteristic of natural sex transformation. Biochem Biophys Res Commun 2003;303:452–457. 45. Kim I, Saunders TL, Morisson SJ. Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem cells. Cell 2007;130:470–483. 46. Robinson LLL, Gaskell TL, Saunders PTK, Anderson RA. Germ cell specific expression of c-kit in the human fetal gonad. Mol Hum Reprod 2001;7:845–852. 47. de Jong J, Stoop H, Gillis AJM, Hersmus R, Van Gurp RJHLM, Van de Geijn G-JM,et al. Further characterization of the first seminoma cell line TCam-2. Genes Chromosomes Cancer 2007;47:185–196. 48. Fogh J. Cultivation, characterization, and identification of human tumor cells with emphasis on kidney, testis, and bladder tumors. Natl Cancer Inst Monogr 1978;49:5–9.

134 SOX17 AND SOX2 IN GERM AND STEM CELLS

49. Andrews PW, Damjanov I, Simon D, Banting GS, Carlin C, Dracopoli NC, et al. Pluripotent embryonal carcinoma clones derived from the human teratocarcinoma cell line Tera-2. Differentiation in vivo and in vitro. Lab Invest 1984;50:147–162. 50. Wang N, Trend B, Bronson DL, Fraley EE. Nonrandom abnor-malities in chromosome 1 in human testicular cancers. Cancer Res 1980;40:796–802. 51. Mavilio F, Simeone A, Boncinelli E, Andrews PW. Activation of four homeobox gene clusters in human embryonal carcinoma cells induced to differentiate by retinoic acid. Differentiation 1988;37:73–79. 52. Teshima S, Shimosato Y, Hirohashi S, Tome Y, Hayashi I, Kanazawa H, et al. Four new human germ cell tumor cell lines. Lab Invest 1988;59:328–336. 53. Burger H, Nooter K, Boersma AW, van Wingerden KE, Looijenga LH, Jochemsen AG, et al. Distinct p53-independent apoptotic cell death signalling pathways in testicular germ cell tumour cell lines. Int J Cancer 1999;81:620–628. 54. de Jong J, Stoop H, Gillis AJM, Van Gurp RJHLM, Van Drunen E, Beverloo HB, et al. JKT-1 is not a human seminoma cell line. Int J Androl 2007;30:350–365. 55. Looijenga LHJ, Gillis AJM, Stoop H, Hersmus R, Oosterhuis JW. Chromosomes and expression in human testicular germ cell tumors: insight in origin and pathogenesis. N Y Acad Sci 2008;1120:187–214. 56. Santagata S, Ligon KL, Hornick JL. Embryonic stem cell transcription factor signatures in the diagnosis of primary and metastatic germ cell tumors. Am J Surg Pathol 2007; 31:836–845. 57. Phi JH, Park SH, Paek SH, Kim SK, Lee YJ, Park CK, et al. Expression of Sox2 in mature and immature teratomas of central nervous system. Mod Pathol 2007;20:742–748. 58. Wood HB, Episkopou V. Comparative expression of the mouse Sox1, Sox2 and Sox3 genes from pre-gastrulation to early somite stages. Mech Dev 1999;86:197–201. 59. Kiernan AE, Pelling AL, Leung KK, Tang AS, Bell DM, Tease C, et al. Sox2 is required for sensory organ development in the mammalian inner ear. Nature 2005;434:1031–1035. 60. Bani-Yaghoub M, Tremblay RG, Lei JX, Zhang D, Zurakowski B, Sandhu JK, et al. Role of Sox2 in the development of the mouse neocortex. Dev Biol 2006;295:52–66. 61. Collignon J, Sockanathan S, Hacker A, Cohen-Tannoudji M, Norris D, Rastan S, et al. A comparison of the properties of Sox-3 with Sry and two related genes, Sox-1 and Sox-2. Development 1996;122:509–520. 62. Payen E, Pailhoux E, Gianquinto L, Hayes H, Le Pennec N, Bezard J, et al. The ovine SOX2 gene: sequence, chromosomal localization and gonadal expression. Gene 1997;189:143– 147. 63. Western P, Maldonado-Saldivia J, van den Bergen J, Hajkova P, Saitou M, Barton S, et al. Analysis of Esg1 expression in pluripotent cells and the germline reveals similarities with Oct4 and Sox2 and differences between human pluripotent cell lines. Stem Cells 2005;23:1436–1442. 64. Yabuta Y, Kurimoto K, Ohinata Y, Seki Y, Saitou M. Gene expression dynamics during germline specification in mice identified by quantitative single-cell gene expression profiling. Biol Reprod 2006;75:705–716. 65. Kanai-Azuma M, Kanai Y, Gad JM, Tajima Y, Taya C, Kurohmaru M, et al. Depletion of definitive gut endoderm in Sox17-null mutant mice. Development 2002;129:2367–2379. 66. Takao Y, Yokota T, Koide H. Beta-catenin up-regulates Nanog expression through interaction with Oct-3/4 in embryonic stem cells. Biochem Biophys Res Commun 2006;353:699–705. 67. Sinner D, Rankin S, Lee M, Zorn AM. Sox17 and beta-catenin cooperate to regulate the transcription of endodermal genes. Development 2004;131:3069–3080.

135

A:IM>KØ×

&G8@68??946GBE4F4ABI8?7<4:ABFG<6 @4E>8E9BE84E?L@4?<:A4AG:8E@68??F

BHEA4?B9#4G;B?B:LÒÛÖ ÔÓ ßÔ ÒÚÚÙ

4AF&GBBC E<878@4AABA86>8E 8EG4AI4A788<=A 7 CHAPTER 10

Abstract Carcinoma in situ (CIS) of the testis is the pre-invasive stage of type II testicular germ cell tumours (TGCTs) of adolescents and adults. These tumours are the most frequently diagnosed cancer in Caucasian adolescents and young adults. In dysgenetic gonads, the precursor of type II GCTs can be either CIS or a lesion known as gonadoblastoma (GB). CIS/GB originates from a primordial germ cell (PGC)/gonocyte, ie an embryonic cell. CIS can be cured by local low-dose irradiation, with limited side effects on hormonal function. Therefore, strategies for early diagnosis of CIS are essential. Various markers are informative to diagnose CIS in adult testis by immunohistochemistry, including c-KIT, PLAP, AP- 2y , NANOG, and POU5F1 (OCT3/4). OCT3/4 is the most informative and consistent in presence and expression level, resulting in intense nuclear staining. In the case of maturational delay of germ cells, frequently present in gonads of individuals at risk for type II (T)GCTs, use of these markers can result in overdiagnosis of malignant germ cells. This demonstrates the need for a more specific diagnostic marker to distinguish malignant germ cells from germ cells showing maturation delay. Here we report the novel finding that immunohistochemical detection of stem cell factor (SCF), the c-KIT ligand, is informative in this context. This was demonstrated in over 400 cases of normal (fetal, neonatal, infantile, and adult) and pathological gonads, as well as TGCT- derived cell lines, specifically in cases of CIS and GB. Both membrane-bound and soluble SCF were expressed, suggestive of an autocrine loop. SCF immunohistochemistry can be a valuable diagnostic tool, in addition to OCT3/4, to screen for precursor lesions of TGCTs, especially in patients with germ cell maturation delay.

138 SCF MARKS MALIGNANT GERM CELLS Introduction Type II testicular germ cell tumours (TGCTs) are the most frequent malignancy among Caucasian men aged 20–45 years [1,2], with rising incidence in most Western countries [3,4]. TGCTs originate from a common precursor, known as carcinoma in situ (CIS) or intratubular germ cell neoplasia unclassified (ITGCNU) [2,5]. Dysgenetic gonads show a higher incidence of these tumours [2], of which gonadoblastoma (GB), or CIS, is the precursor [6–8]. CIS/GB cells resemble fetal germ cells [primordial germ cells (PGCs) or gonocytes] by morphology [9] and an erased pattern of genomic imprinting [10]. They also express embryonic markers, including PLAP [11,12], POU5F1 (OCT3/4) [13], NANOG [14], AP-2y [15,16], and c- KIT [17]. These are found in normal fetal germ cells and are lost upon further maturation [11,12,14,16]. Germ cell maturation delay can lead to overdiagnosis and overtreatment [18–20], as in cases of cryptorchidism and disorders of sex development (DSD), ie hypovirilization or gonadal dysgenesis [21–25] (for reviews see refs 18 and 26). [Correction made here after initial online publication]. Although criteria have been developed to distinguish malignant germ cells from cells showing delayed maturation only in cases of hypovirilization [27], reliable diagnostic tools are lacking. Therefore a specific marker for malignant germ cells would be informative.

Normal germ cell lineage development depends on the c-KIT–stem cell factor (SCF) pathway [28]. In mice, c-KIT (the receptor) is present in PGCs, gonocytes, spermatogonia, spermatocytes, round spermatids, and Leydig cells [29–33]. c-KIT signalling is also important for human spermatogenesis, demonstrated by defects in this pathway (for a review see ref 34). However, the pattern of expression is different between mice and humans. In the latter, c-KIT expression is low and is not detected in adult testis, unless frozen tissue and a sensitive detection method are used [17,35–37]. This makes c-KIT a diagnostic marker for CIS in adults on formalin-fixed, paraffin-embedded (FFPE) material [11,17]. However, false-negative findings have been reported [38,39].In adult testes, the c-KIT ligand, stem cell factor (SCF), is produced by Sertoli cells [34,40–42], under the control of follicle-stimulating hormone (FSH) [43]. Due to alternative RNA splicing and post-translational processing, two forms of SCF exist: a membrane- bound and a soluble protein [44,45]. Membrane-bound SCF is the most efficient in establishing and maintaining germ cells [46–50], while soluble SCF activates c- KIT on Leydig cells to induce testosterone production [51].

We investigated the presence of SCF in a series of normal and pathological gonads. SCF is virtually absent in normal testis, while all cases of CIS and GB were positive. No SCF was found to be related to germ cells with proven maturational delay, unlike the other known CIS/GB markers. This demonstrates the additional

139 CHAPTER 10 value of immunohistochemical detection of SCF to identify malignant germ cells, especially in the case of germ cell maturation delay.

Materials and methods Use of tissues for scientific purposes was approved by an institutional review board (MEC 02.981 and CCR2041). Samples were used according to the ‘Code for Proper Secondary Use of Human Tissue in The Netherlands’, from the Dutch Federation of Medical Scientific Societies (FMWV) (version 2002).

Pathological tissues Tissue samples were collected as described before [52] and tumours were diagnosed according to WHO standards [5]. Male gonads, obtained after spontaneous or induced abortions (gestational age 15–40 weeks) or from premature and term neonates that died shortly after birth, were included as previously described [11]. Table 1. Gonadal tissues n Positive % SCF immunohistochemical staining data of FFPE Fetal and infantile testicular tissue material of gonads of fetal, Fetal gonads 26 2 8 infantile, and hypovirilized Infantile biopsies 7 0 0 patients, as well as CIS, Infantile gonads 20 0 0 TGCTs, and gonadoblasto- Hypovirilization patients 39 5 13 mas. Note the higher percentage of SCF-positive Testicular tissue adults cases when comparing Biopsies 23 3 13 total tissue sections with Total tissue sections TGCT TMA, due to heterogeneity CIS 81 81 100 Seminoma 68 64 94 Non-seminoma 72 47 65 Embryonal carcinoma 25 11 44 Yolk sac tumour 23 20 87 Choriocarcinoma 5 4 80 Teratoma 19 12 63 TMA CIS 43 43 100 Seminoma 50 50 100 Non-seminoma 121 54 45 Embryonal carcinoma 47 14 30 Yolk sac tumour 30 22 73 Choriocarcinoma 17 6 35 Teratoma 27 12 44

Gonadoblastoma 13 13 100

Furthermore, gonads of complete (CAIS) or partial androgen insensitivity syndrome (PAIS), dysgenetic gonads [27], and open surgical testicular biopsies from adult infertile patients as well as gonadal biopsies and complete orchidectomy samples from routine paediatric urological practice were included (Table 1).

140 SCF MARKS MALIGNANT GERM CELLS For a quick scan, tissue microarrays (TMAs) were used, including seminoma, CIS, embryonal carcinoma, yolk sac tumour, choriocarcinoma, and teratoma (see Table 1).

Cell lines Cytospins were prepared from type II (T)GCT-derived cell lines, cultured according to standard procedures, including the embryonal carcinoma cell lines Tera1 [53], 2102Ep [54], NT2 [55], NCCIT [56], and the seminoma-derived cell line TCam-2 [57–59].

Immunohistochemistry The antibodies used have been described before [11,12]. The SCF antibody is a goat polyclonal (N19; Santa Cruz Biotechnology; 1 : 500 to 1 : 350), detecting both membrane-bound and soluble human SCF. Its specificity has been demonstrated previously [60]. Staining was performed as described previously [61]: tissue sections were incubated, after antigen retrieval [62], with primary antibody overnight at 40C (PLAP, c-KIT, SCF) or for 2 h at room temperature (OCT3/4). Frozen sections and cytospins of cell lines were incubated for 1 h at room temperature with the antibody for SCF and c-KIT after acetone fixation. For negative controls, the primary antibody was omitted, resulting in complete absence of signal.

Double staining was performed as described previously [27]. Immunofluorescence was carried out on acetone-fixed and paraformaldehyde- treated frozen sections of normal testicular parenchyma by using the secondary biotinylated antibody before visualization with avidin-CY3 and counterstaining with Vectashield (Vector)/DAPI (Sigma Chemical Co, St Louis, MO, USA).

Quantitative RT-PCR for c-KIT and SCF RNA isolation, cDNA synthesis, and quantitative RT-PCR for c-KIT and SCF were performed essentially as described previously [63–65]. The primers, amplifying both membrane-bound and soluble SCF, were SCF-f: 5’- AGCCAAGTCTTACAAGGGCA-3’ and SCF-r: 5’-TAAATGAGACCCAAGTCCCG-3’. Mastermix from Finnzyme was used (Bioke, The Netherlands). In total, 79 seminomas, 43 non-seminomas, 19 normal testes, 14 CIS-containing testicular parenchyma cases, and 7 peripheral blood samples were analysed. Statistical analysis was performed using SPSS v11.5 software (SPSS Inc, Chicago, IL, USA).

141 CHAPTER 10 RT-PCR for mRNA analysis of the membrane-bound and soluble SCF RT-PCR was performed to analyse the presence of membrane-bound and soluble SCF, using SCF-f: 5’-ATGACCTTGTGGAGTGCGTC-3’ and SCF-r2: 5’ - ACTTGGCTGTCTCTTCTTCCAG-3’ (Figure 6A). Amplification was performed using Platinum Taq DNA polymerase High Fidelity (Invitrogen) for 2 min at 950C, followed by 38 cycles of 940C for 30 s, 580C for 30 s, and 680C for 45 s. A boost PCR, identical to the first, was done for 20 additional cycles. miRNA expression analysis RNA isolation and analysis for the expression of known miRNAs with the mirVana miRNA Isolation Kit (Ambion, Austin, TX, USA) and TaqMan MicroRNA Assays Human Panel (Applied Biosystems, Foster City, CA, USA) were carried out according to the manufacturers’ protocol, as previously reported [52].

Results To study the suitability of immunohistochemical detection of SCF, a series of well-characterized histological samples was evaluated. Characteristics of the cases are summarized in Table 1 and the observations are discussed below for the different subgroups.

Normal adult testes Fluorescent and non-fluorescent detection of immunostaining on frozen normal adult testicular tissue (n = 3) showed positive staining for c-KIT in spermatogonia and Leydig cells (Figures 1A and 1B), as reported earlier [35]. Fluorescent visualization of SCF showed diffuse staining in the seminiferous tubules and Leydig cells (Figure 1C), while SCF was not seen with non-fluorescent detection (Figure 1D). c-KIT and SCF were not detectable on FFPE tissue of the same samples, irrespective of pretreatment and visualization protocols (data not shown). All the immunohistochemistry presented in the following sections is on FFPE material, representative for routine pathology.

Fetal gonads SCF immunohistochemistry showed that only two out of 26 fetal gonads had a few positive cells [see Figures 1E (negative) and 1F (positive)]; these were from an 18-week-old male trisomy 21 fetus without further abnormalities and a 16-week- old male fetus with bicephaly. Trisomy 21 is indeed a risk factor for type II TGCTs [66], associated with maturational delay of germ cells [19]. All the others were negative for SCF, including cases of ages similar to those of the positive cases.

142 SCF MARKS MALIGNANT GERM CELLS old old bright (B) fluorescent detection detection fluorescent (C) the different histological histological different the fluorescent detection and and detection fluorescent aining with aining (A) (A) two elements of mature teratoma teratoma mature elements of two (L, M) M) (L, c). SCF immunost c). SCF demonstrates heterogeneity in heterogeneity demonstrates ult testicular ult testicular tissue: reveals a few SCF-positive cells in a gonad of an 18-week- an of a gonad cells in SCF-positive a few reveals fluorescent detection of SCF, being negative in the latter. SCF SCF latter. in the negative being SCF, of detection fluorescent (G–M) (G–M) (F) non- choriocarcinoma; (K) choriocarcinoma; (D) on frozen normal ad normal frozen on nia (Sg) and cells Leydig (L nia (Sg) (C, D) D) (C, yolk sac tumour; tumour; sac yolk (J) and SCF and SCF (A, B) embryonal carcinoma; carcinoma; embryonal (I) (I) seminoma; seminoma; negative in a fetal male gonad of 15 weeks’ gestational age and age gestational 15 weeks’ of gonad male a fetal in negative (E) (G, H) (G, H) Immunohistochemistry for c-KIT c-KIT for Immunohistochemistry Figure 1. spermatogo indicate positive c-KIT. Arrows of detection field (showing diffuselystaining positive in the tubules Leydig and and cells) is staining invasive TGCTs of SCF for Immunohistochemistry 21. with trisomy male fetus subtypes:

143 CHAPTER 10 Invasive TGCTs All TGCTs represented in TMA showed a heterogeneous staining pattern for SCF (Table 1), confirmed on total tissue section (see Figures 1G-1M and Table 1). The majority showed some positive staining; seminomas heterogeneously, both in number of positive cells and in signal intensity (Figures 1G and 1H). Non- seminomas showed different patterns in histological subgroups (Figures 1I–1M) (positive: 44% of embryonal carcinoma, 87% of yolk sac tumours, and 80% of choriocarcinomas). In 63% of the teratomas, positive cells were found in various types of immature and mature tissue (eg stratified epithelium, cytoplasm of cylindrical cells, and subepithelial cells with cylindrical differentiation, Figures 1L and 1M).

Pre-invasive lesions: CIS and GB All CIS samples, both adjacent to seminoma and adjacent to non-seminoma, represented in TMA and total sections, were positive for SCF (see Table 1 and Figure 2A). This was also found for CIS before the development of invasiveness (n = 3) (Figure 2B). Double staining (OCT3/4 and SCF) confirmed the identity of CIS cells (Figures 2C and 2D). Like CIS, all GBs (n = 13) and intratubular seminomas were positive for SCF (Figures 2E and 2F).

Figure 2. Immunohistochemical staining for SCF (red). (A) CIS adjacent to an invasive seminoma as well as (B) CIS adjacent to normal testicular tissue. (C, D) double staining for OCT3/4 (red) and SCF (blue) on CIS showing copresence; (E) gonadoblastoma and (F) intratubular seminoma are both positive for SCF as well.

144 SCF MARKS MALIGNANT GERM CELLS Hypovirilization and infantile patients A series of patients with various forms of hypovirilization was investigated, all known to have a certain delay in germ cell maturation, previously characterized using PLAP, c-KIT, and OCT3/4 [27] (Figures 3A–3C). No CIS was diagnosed based on the WHO criteria and no SCF was detected using immunohistochemistry (Figure 3D). In contrast, we reported three cases of hypovirilization with CIS, based on patient age, staining, and distribution of OCT3/4-positive cells [27]. These cases were indeed positive for SCF (Figures 3E and 3F), restricted to the CIS- containing seminiferous tubules.

Figure 3. Immunohistochemistry on parallel sections of gonadal tissue of a DSD patient with hypovirilization without CIS. (A) OCT3/4. Inset H&E staining (B) PLAP, (C) c-KIT, and (D) SCF. There is colocalization of all markers except SCF, which is negative (D). In contrast, SCF is detected in the gonads of two DSD patients with CIS (E, F), identified by OCT3/4 staining (inset). (G)Staining of the gonadal biopsies of an infantile DSD patient demonstrating that SCF-positive cells and OCT3/4-positive cells (indicated by circles in H) are seen in the same areas. See text for details.

145 CHAPTER 10 Two additional DSD patients were included: the first case was a girl (46,XY) with ambiguous genitalia (biopsy revealed a streak gonad and a dysgenetic testis). OCT3/4-positive cells at the basal lamina of the seminiferous tubules were identified in the testis, although not judged as classical CIS. This case also showed SCF staining, although not of the same cells (Figures 3G and 3H). The second case was a 1-year-old DSD patient (46,XY) with ambiguous genitalia (severe hypospadias, bilateral intra-abdominal gonads, and a vagina). OCT3/4-positive cells were found in the lumen of some seminiferous tubules, although in addition there were some on the basal lamina (data not shown). Indeed, again, a limited number of SCF-positive germ cells were identified. Both patients are under active and close clinical surveillance. A series of testicular biopsies (n = 7) and testes obtained by orchidectomy (n = 20), taken in routine paediatric urological practice, was investigated. All samples were negative for SCF staining (data not shown).

Relationship between SCF staining and c-KIT Based on parallel staining for SCF and c-KIT, no relationship was observed between these markers (data not shown), with the exception of CIS/GB and seminomas. Two additional seminomas with a genomic c-KIT amplification were investigated [65], showing the same finding. Of the samples investigated (see above), three cases have a c-KIT exon 17 mutation (two codon 816 and one 822) and 15 are wild type. Again, no relationship with SCF was observed. mRNA analysis of SCF and c-KIT Seminomas showed variable levels of c-KIT (Figure 4A), which was overall lower in non-seminomas (Figure 4B). Furthermore, expression of c-KIT was low in normal testicular parenchyma, and in parenchyma containing various amounts of CIS, with-out a relationship with the percentage of CIS. No c-KIT expression was found in peripheral blood, whereas SCF mRNA was present (Figure 4B). SCF expression was heterogeneous in seminomas, with high expression in four cases, which showed c-KIT expression that was significantly lower (indicated by asterisks in Figure 4A, Mann–Whitney test, p = 0.027). No difference in SCF immunohistochemistry, which is non-quantitative, was found in these four cases compared with the others. Non-seminomas showed heterogeneous SCF mRNA expression, in accordance with immunohistochemistry. Normal testicular parenchyma expressed SCF mRNA, as did the samples with various amounts of CIS, without a relationship between these two parameters (Figure 4B). Overall, there was no significant correlation between SCF and c-KIT mRNA levels (Spearman’s rank correlation, p = 0.35). When the different subgroups were analysed, only the non-seminomas showed a significant correlation between c- KIT and SCF mRNA expression (Spearman rank correlation coefficient: + 0.557, p < 0.001). Although of interest, this was not investigated further in this study.

146 SCF MARKS MALIGNANT GERM CELLS To discriminate between the mRNA expression of membrane-bound and soluble SCF, RT-PCR (Figure 5A) was performed on normal testis, CIS, seminomas, and non-seminomas. In general, both forms of SCF were found (Figure 5B), although one seminoma and one non-seminoma were negative for both forms, possibly due to heterogeneity of SCF expression by immunohistochemistry or differences between FFPE and frozen material.

Figure 4. Quantitative RT-PCR for c-KIT and SCF expression (A) seminomas (n = 79) and (B) non- seminomas (n = 43), normal testis (n = 19), testes with different amounts of CIS (n = 14), and peripheral blood leukocytes (PBL) (n = 7). All samples are displayed sorted by their SCF expression level. Seminoma samples with the highest expression of SCF and low expression of c-KIT are indicated (asterisks in A). Expression of c-KIT and SCF mRNA is normalized based on HPRT mRNA expression that was determined in parallel as a control. Note the logarithmic scales on the y-axis

TGCT cell lines, c-KIT, and SCF Whereas there are no cell lines for CIS or GB, cell lines representative of embryonal carcinoma and, most recently, also seminoma are available. With the exception of NCCIT, being negative for both SCF and c-KIT protein, all the other cell lines showed at least some positive staining for SCF and c-KIT (Figure 6A). Only a minority of positive cells were found in NT2, 2102EP, and Tera-1, while the majority of TCam-2 cells were positive. Because of technical limitations, we could not prove that c-KIT and SCF are present on the same cells, although this is most likely. Expression of c-KIT and SCF was confirmed on the mRNA level by quantitative PCR, which showed that all the cell lines, including NCCIT, express c- KIT and SCF at varying levels (Figure 6B), again without a relation-ship between the two. Like the invasive tumours, the cell lines also expressed overall both

147 CHAPTER 10 membrane-bound and soluble SCF (Tera-1 and TCam-2 the lowest levels of membrane-bound form) (Figure 6C). The absence of a relationship between mRNA and protein could be due to specific microRNAs. However, this was not observed for three miRNAs known to interact with c-KIT mRNA, ie hsa-miR-146, - 221, and -222 [67] (Figure 6D).

Discussion Clinical diagnosis of CIS in adolescents and adults is based on the detection of markers such as PLAP, c-KIT, and OCT3/4 [11]. These markers are specifically found in PGCs/gonocytes, but not at later developmental stages of germ cells. This hampers CIS diagnosis at an early age and in cases of germ cell maturation delay [9]. This may result in overdiagnosis, possibly preventable in cases of undervirilization based on additional criteria for the diagnosis of CIS. These include the presence of OCT3/4 and TSPY-positive cells on the basal lamina of seminiferous tubules, showing clonal expansion, especially at ages above 3 months [27]. However, the identification of a marker that discriminates malignant germ cells from germ cells showing delayed maturation would be of clinical diagnostic value.

SCF–c-KIT signalling is important for PGC migration, survival, proliferation, and maturation [41,68–71]. Activating mutations of c-KIT have been reported in type II TGCTs, predominantly in bilateral cases [72–74], and specific amplification of the c-KIT locus has been described [65]. We therefore studied the presence of SCF during normal and pathological gonadal development, including type II TGCTs. SCF is possibly involved in the survival of the precursors of type II TGCTs, CIS and GB, in their specific micro-environment.

CIS and GB were investigated using immunohistochemistry for SCF, alongside the established markers OCT3/4, PLAP, and c-KIT. Interestingly, all precursor lesions stained positive for SCF, independently of the presence of an invasive component. In contrast, a heterogeneous pattern was found in invasive tumours. Testes of fetuses and newborns showed no SCF staining, with the exception of two cases, one of which belonged to a risk group for type II GCT. Moreover, gonads of DSD patients were investigated; those with no (or low) tumour risk lacked SCF staining, while cases previously identified with a high risk stained positively for SCF.

148 SCF MARKS MALIGNANT GERM CELLS

Figure 5. (A) Analysis of SCF variant RNA expression in type II TGCTs. Location of primers f and r2 (arrows) used to detect membrane-bound and soluble SCF by RT-PCR are indicated. (B) SCF RT-PCR demonstrating the expression of soluble and membrane-bound SCF transcripts in normal testis, testicular tissue containing CIS, seminoma (SE), and non-seminoma (EC = embryonal carcinoma; YS = yolk sac tumour; TE = teratoma; MQ = MilliQ negative control). HPRT expression was detected to check the quality of the cDNA

How SCF is involved in the pathogenesis of type II (T)GCTs remains to be elucidated. No relationship was found between SCF and c-KIT expression or mutations, and both the membrane-bound and the soluble form of SCF were present in most tumours and cell lines. The level of mRNA expression did not parallel the protein findings, which could not be explained by the limited number of miRNAs investigated. An alternative explanation might be accumulation of SCF in the CIS containing seminiferous tubules, possibly due to the presence of c- KIT. This might be in line with the absence of SCF on some individual CIS cells, despite its presence in (possibly interconnected) adjacent seminiferous tubules. However, the observations in cell lines support the model of an autocrine loop, a deviation from the physiological paracrine interaction during normal germ cell development and spermatogenesis. This may explain the cytoplasmic SCF

149 CHAPTER 10 staining of CIS and seminoma cells, as reported before [60], although internalization might occur. If an autocrine loop is involved, it is probably predominantly of relevance in the early stages of pathogenesis.

Figure 6. Expression of SCF and c-KIT studied in a panel of embryonal carcinoma (NCCIT, NT-2, 2102 EP, and Tera-1) and seminoma (TCam-2) cell lines. (A) Immunohistochemistry; (B) quantitative RT-PCR; (C) RT-PCR discriminating soluble and membrane-bound SCF; (D) expression analysis of three miRNAs (miR 146, 221 and 222) reported to target c-KIT. With the exception of NCCIT, all the embryonal carcinoma cell lines showed a certain percentage of c-KIT and SCF-positive cells, while almost all cells of the seminoma cell line TCam-2 were positive. All the cell lines express SCF and c-KIT mRNA. There is no relationship between SCF and c-KIT protein expression on one side versus the mRNA levels, or the presence of both the membrane-bound and the soluble SCF form. Moreover, no relationship between the expression level of three (c-KIT targeting) miRNAs and the level of c-KIT mRNA or protein was identified. (E) Proposed model for the role of SCF–c-KIT signalling during the development and progression of type II TGCTs. In the initial, embryonal pre-malignant stage, SCF signalling is paracrine, produced by the Sertoli cells (I), possibly with higher SCF levels than in the normal embryonal situation (IV). Upon progression to CIS, the cells are dependent on SCF (illustrated by our observation that all CIS are positive for SCF) and an autocrine stimulation loop has developed (II). Upon progression to invasiveness, the cells become independent of SCF, resulting in heterogeneous expression patterns of SCF, eg a seminoma with diminished expression of SCF (III). During normal embryonal to adult development, the germ cells are dependent on a low level of paracrine SCF stimulation, only detectable with sensitive methods (IV, V, and Figure 1C)

The cell lines can be informative in this context, especially TCam-2, which is representative of seminoma and has significant similarities to CIS [58]. Reduced viability of TCam-2 cells after siRNA-induced down-regulation of c-KIT has indeed been demonstrated recently, suggestive of an autocrine stimulation loop [59]. It is not clear whether bovine SCF, capable of activating human c-KIT, present in the fetal calf serum (FCS) of the cell culture medium, may also play a role [75,76]. 150 SCF MARKS MALIGNANT GERM CELLS Growth stimulation by SCF of non-seminoma cell lines was seen when grown in the absence, but not when grown in the presence, of FCS [35], which might indeed indicate interference of bovine SCF and that the autocrine loop may not be present in non-seminomas. However, this needs further investigation. The exact mechanism by which SCF is involved in the pathogenesis of type II TGCTs remains to be elucidated. The most interesting model is a paracrine to an autocrine switch during early development and the subsequent independence of this pathway upon further progression (Figure 6E). This altered bioavailability of SCF (it appears to accumulate on CIS, so less SCF may be available for activating Leydig cells) may lead to reduced testosterone production, resulting in the activation of a feedback loop to produce LH [77]. This could explain the occurrence of Leydig cell hyperplasia, as is seen in about half of the cases of testicular parenchyma containing CIS [78,79].

Our study demonstrates that SCF, in addition to OCT3/4, can be used as a marker allowing the distinction of malignant germ cells from those with delayed maturation, particularly in gonads at risk for TGCT and showing germ cell maturation delay. The diagnostic possibilities of detecting CIS cells in young individuals and patients with various forms of DSD will benefit significantly from this observation. Moreover, the findings support the model that a switch from paracrine to autocrine SCF–c-KIT signalling could be relevant in the early pathogenesis of type II GCTs.

Acknowledgements We thank Ewout Steyerberg, Department of Public Health, Erasmus MC University Medical Center Rotterdam, for help with the statistical analysis of the data. FH is supported by Wilhelm Sander-Stiftung and Werner Otto Stiftung, Germany; GJMvdG and MdB are supported by a grant from the Dutch Cancer Society (KWF, grant No 2004-3123); and MC is supported by the Research Foundation Flanders.

References 1. Schmoll HJ, Souchon R, Krege S, Albers P, Beyer J, Kollmannsberger C, et al. European consensus on diagnosis and treatment of germ cell cancer: a report of the European Germ Cell Cancer Consensus Group (EGCCCG). Ann Oncol 2004;15:1377–1399. 2. Oosterhuis JW, Looijenga LH. Testicular germ-cell tumours in a broader perspective. Nature Rev Cancer 2005;5:210–222. 3. Purdue MP, Devesa SS, Sigurdson AJ, McGlynn KA. International patterns and trends in testis cancer incidence. Int J Cancer 2005;115:822–827. 4. Shah MN, Devesa SS, Zhu K, McGlynn KA. Trends in testicular germ cell tumours by ethnic group in the United States. Int J Androl 2007;30:206–213; discussion 213-204. 5. Woodward PJ, Heidenreich A, Looijenga LHJ, Oosterhuis JW, McLeod DG, Møller H, et al. World Health Organization classification of tumours: tumours of the testis and paratesticular tissue. In Pathology and Genetics of Tumours of the Urinary System and 151 CHAPTER 10

Male Genital Organs, Eble JNSG, Epstein JI, Sesterhenn IA (eds). IARC Press: Lyon, 2004; 217– 249. 6. Robboy SJ, Miller T, Donahoe PK, Jahre C, Welch WR, Haseltine FP, et al. Dysgenesis of testicular and streak gonads in the syndrome of mixed gonadal dysgenesis: perspective derived from a clinicopathologic analysis of twenty-one cases. Hum Pathol 1982;13:700– 716. 7. Verp MS, Simpson JL. Abnormal sexual differentiation and neoplasia. Cancer Genet Cytogenet 1987;25:191–218. 8. Cools M, Stoop H, Kersemaekers AM, Drop SL, Wolffenbuttel KP, Bourguignon JP, et al. Gonadoblastoma arising in undifferentiated gonadal tissue within dysgenetic gonads. J Clin Endocrinol Metab 2006;91:2404–2413. 9. Rajpert-De Meyts E, Bartkova J, Samson M, Hoei-Hansen CE, Frydelund-Larsen L, Bartek J, et al. The emerging phenotype of the testicular carcinoma in situ germ cell. Apmis 2003;111:267–278; discussion 278–279. 10. van Gurp RJ, Oosterhuis JW, Kalscheuer V, Mariman EC, Looijenga LH. Biallelic expression of the H19 and IGF2 genes in human testicular germ cell tumors. J Natl Cancer Inst 1994;86:1070–1075. 11. Honecker F, Stoop H, de Krijger RR, Chris Lau YF, Bokemeyer C, Looijenga LH. Pathobiologicalimplications of the expression of markers of testicular carcinoma in situ by fetal germ cells. J Pathol 2004;203:849–857. 12. Stoop H, Honecker F, Cools M, de Krijger R, Bokemeyer C, Looijenga LH. Differentiation and development of human female germ cells during prenatal gonadogenesis: an immunohistochemical study. Hum Reprod 2005; 20:1466–1476. 13. Looijenga LH, Stoop H, de Leeuw HP, de Gouveia Brazao CA, Gillis AJ, van Roozendaal KE, et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 2003;63:2244–2250. 14. Hoei-Hansen CE, Almstrup K, Nielsen JE, Brask Sonne S, Graem N, Skakkebæk NE, et al. Stem cell pluripotency factor NANOG is expressed in human fetal gonocytes, testicular carcinoma in situ and germ cell tumours. Histopathology 2005;47:48–56. 15. Hoei-Hansen CE, Nielsen JE, Almstrup K, Sonne SB, Graem N, Skakkebæk NE, et al. Transcription factor AP-2gamma is a developmentally regulated marker of testicular carcinoma in situ and germ cell tumors. Clin Cancer Res 2004;10:8521–8530. 16. Pauls K, Jager R, Weber S, Wardelmann E, Koch A, Buttner R, et al. Transcription factor AP- 2gamma, a novel marker of gonocytes and seminomatous germ cell tumors. Int J Cancer 2005;115:470–477. 17. Rajpert-De Meyts E, Skakkebæk NE. Expression of the c-kit protein product in carcinoma-in- situ and invasive testicular germ cell tumours. Int J Androl 1994;17:85–92. 18. Cools M, Drop SL, Wolffenbuttel KP, Oosterhuis JW, Looijenga LH. Germ cell tumors in the intersex gonad: old paths, new directions, moving frontiers. Endocr Rev 2006;27:468–484. 19. Cools M, Honecker F, Stoop H, Veltman JD, de Krijger RR, Steyerberg E, et al. Maturation delay of germ cells in fetuses with trisomy 21 results in increased risk for the development of testicular germ cell tumors. Hum Pathol 2006;37:101–111. 20. Rajpert-De Meyts E, Jorgensen N, Brondum-Nielsen K, Muller J, Skakkebæk NE. Developmental arrest of germ cells in the pathogenesis of germ cell neoplasia. Apmis 1998;106:198–204; discussion 204–206. 21. Muller J. Morphometry and histology of gonads from twelve children and adolescents with the androgen insensitivity (testicular feminization) syndrome. J Clin Endocrinol Metab 1984;59:785–789. 22. Slowikowska-Hilczer J, Szarras-Czapnik M, Kula K. Testicular pathology in 46,XY dysgenetic male pseudohermaphroditism: an approach to pathogenesis of testis cancer. J Androl 2001;22:781–792.

152 SCF MARKS MALIGNANT GERM CELLS

23. Slowikowska-Hilczer J, Romer TE, Kula K. Neoplastic potential of germ cells in relation to disturbances of gonadal organogenesis and changes in karyotype. J Androl 2003;24:270– 278. 24. Jacobsen GK, Henriques UV. A fetal testis with intratubular germ cell neoplasia (ITGCN). Mod Pathol 1992;5:547–549. 25. Cassio A, Cacciari E, D’Errico A, Balsamo A, Grigioni FW, Pascucci MG, et al. Incidence of intratubular germ cell neoplasia in androgen insensitivity syndrome. Acta Endocrinol (Copenhagen) 1990;123:416–422. 26. Looijenga LH, Hersmus R, Oosterhuis JW, Cools M, Drop SL, Wolffenbuttel KP. Tumor risk in disorders of sex development (DSD). Best Pract Res Clin Endocrinol Metab 2007;21:480– 495. 27. Cools M, van Aerde K, Kersemaekers AM, Boter M, Drop SL, Wolffenbuttel KP, et al. Morphological and immunohistochemical differences between gonadal maturation delay and early germ cell neoplasia in patients with undervirilization syndromes. J Clin Endocrinol Metab 2005;90:5295–5303. 28. Ginsburg M, Snow MH, McLaren A. Primordial germ cells in the mouse embryo during gastrulation. Development 1990;110:521–528. 29. Manova K, Nocka K, Besmer P, Bachvarova RF. Gonadal expression of c-kit encoded at the W locus of the mouse. Development 1990;110:1057–1069. 30. Manova K, Bachvarova RF. Expression of c-kit encoded at the W locus of mice in developing embryonic germ cells and presumptive melanoblasts. Dev Biol 1991;146:312–324. 31. Sorrentino V, Giorgi M, Geremia R, Besmer P, Rossi P. Expression of the c-kit proto-oncogene in the murine male germ cells. Oncogene 1991;6:149–151. 32. Yoshinaga K, Nishikawa S, Ogawa M, Hayashi S, Kunisada T, Fujimoto T. Role of c-kit in mouse spermatogenesis: identification of spermatogonia as a specific site of c-kit expression and function. Development 1991;113:689–699. 33. Rossi P, Marziali G, Albanesi C, Charlesworth A, Geremia R, Sorrentino V. A novel c-kit transcript, potentially encoding a truncated receptor, originates within a kit gene intron in mouse spermatids. Dev Biol 1992;152:203–207. 34. Mauduit C, Hamamah S, Benahmed M. Stem cell factor/c-kit system in spermatogenesis. Hum Reprod Update 1999;5:535–545. 35. Bokemeyer C, Kuczyk MA, Dunn T, Serth J, Hartmann K, Jonasson J, et al. Expression of stem-cell factor and its receptor c-kit protein in normal testicular tissue and malignant germ-cell tumours. J Cancer Res Clin Oncol 1996;122:301–306. 36. Strohmeyer T, Reese D, Press M, Ackermann R, Hartmann M, Slamon D. Expression of the c- kit proto-oncogene and its ligand stem cell factor (SCF) in normal and malignant human testicular tissue. J Urol 1995;153:511–515. 37. Natali PG, Nicotra MR, Sures I, Santoro E, Bigotti A, Ullrich A. Expression of c-kit receptor in normal and transformed human nonlymphoid tissues. Cancer Res 1992;52:6139–6143. 38. Rajpert-De Meyts E, Kvist M, Skakkebæk NE. Heterogeneity of expression of immunohistochemical tumour markers in testicular carcinoma in situ: pathogenetic relevance. Virchows Arch 1996;428:133–139. 39. Izquierdo MA, Van der Valk P, Van Ark-Otte J, Rubio G, Germa-Lluch JR, Ueda R, et al. Differential expression of the c-kit protooncogene in germ cell tumours. J Pathol 1995;177:253–258. 40. Manova K, Huang EJ, Angeles M, De Leon V, Sanchez S, Pronovost SM, et al. The expression pattern of the c-kit ligand in gonads of mice supports a role for the c-kit receptor in oocyte growth and in proliferation of spermatogonia. Dev Biol 1993;157:85–99. 41. Hakovirta H, Yan W, Kaleva M, Zhang F, Vanttinen K, Morris PL, et al. Function of stem cell factor as a survival factor of spermatogonia and localization of messenger ribonucleic acid in the rat seminiferous epithelium. Endocrinology 1999;140:1492–1498.

153 CHAPTER 10

42. Sandlow JI, Feng HL, Cohen MB, Sandra A. Expression of c-KIT and its ligand, stem cell factor, in normal and subfertile human testicular tissue. J Androl 1996;17:403–408. 43. Rossi P, Dolci S, Albanesi C, Grimaldi P, Ricca R, Geremia R. Follicle-stimulating hormone induction of steel factor (SLF) mRNA in mouse Sertoli cells and stimulation of DNA synthesis in spermatogonia by soluble SLF. Dev Biol 1993;155:68–74. 44. Flanagan JG, Chan DC, Leder P. Transmembrane form of the kit ligand growth factor is determined by alternative splicing and is missing in the Sld mutant. Cell 1991;64:1025– 1035. 45. Huang EJ, Nocka KH, Buck J, Besmer P. Differential expression and processing of two cell associated forms of the kit-ligand: KL-1 and KL-2. Mol Biol Cell 1992;3:349–362. 46. Godin I, Deed R, Cooke J, Zsebo K, Dexter M, Wylie CC. Effects of the steel gene product on mouse primordial germ cells in culture. Nature 1991;352:807–809. 47. Dolci S, Williams DE, Ernst MK, Resnick JL, Brannan CI, Lock LF, et al. Requirement for mast cell growth factor for primordial germ cell survival in culture. Nature 1991;352:809–811. 48. Matsui Y, Toksoz D, Nishikawa S, Williams D, Zsebo K, Hogan BL. Effect of steel factor and leukaemia inhibitory factor on murine primordial germ cells in culture. Nature 1991;353:750–752. 49. Marziali G, Lazzaro D, Sorrentino V. Binding of germ cells to mutant Sld Sertoli cells is defective and is rescued by expression of the transmembrane form of the c-kit ligand. Dev Biol 1993;157:182–190. 50. Broudy VC. Stem cell factor and hematopoiesis. Blood 1997;90:1345–1364. 51. Yan W, Kero J, Huhtaniemi I, Toppari J. Stem cell factor functions as a survival factor for mature Leydig cells and a growth factor for precursor Leydig cells after ethylene dimethane sulfonate treatment: implication of a role of the stem cell factor/c-Kit system in Leydig cell development. Dev Biol 2000;227:169–182. 52. Gillis AJ, Stoop HJ, Hersmus R, Oosterhuis JW, Sun Y, Chen C, et al. High-throughput microRNAome analysis in human germ cell tumours. J Pathol 2007;213:319–328. 53. Fogh J. Cultivation, characterization, and identification of human tumor cells with emphasis on kidney, testis, and bladder tumors. Natl Cancer Inst Monogr 1978;5–9. 54. Wang N, Trend B, Bronson DL, Fraley EE. Nonrandom abnormalities in chromosome 1 in human testicular cancers. Cancer Res 1980;40:796–802. 55. Andrews PW, Damjanov I, Simon D, Banting GS, Carlin C, Dracopoli NC, et al. Pluripotent embryonal carcinoma clones derived from the human teratocarcinoma cell line Tera-2. Differentiation in vivo and in vitro. Lab Invest 1984;50:147–162. 56. Teshima S, Shimosato Y, Hirohashi S, Tome Y, Hayashi I, Kanazawa H, et al. Four new human germ cell tumor cell lines. Lab Invest 1988;59:328–336. 57. Mizuno Y, Gotoh A, Kamidono S, Kitazawa S. [Establishment and characterization of a new human testicular germ cell tumor cell line (TCam-2)]. Nippon Hinyokika Gakkai Zasshi 1993;84:1211–1218. 58. de Jong J, Stoop H, Gillis AJ, Hersmus R, van Gurp RJ, van de Geijn GJ, et al. Further characterization of the first seminoma cell line TCam-2. Genes Chromosomes Cancer 2008;47:185–196. 59. Goddard NC, McIntyre A, Summersgill B, Gilbert D, Kitazawa S, Shipley J. KIT and RAS signalling pathways in testicular germ cell tumours: new data and a review of the literature. Int J Androl 2007;30:337–348; discussion 349. 60. Gashaw I, Dushaj O, Behr R, Biermann K, Brehm R, Rubben H, et al. Novel germ cell markers characterize testicular seminoma and fetal testis. Mol Hum Reprod 2007; 13:721–727. 61. Stoop H, van Gurp R, de Krijger R, Geurts van Kessel A, Koberle B, Oosterhuis W, et al. Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological implications. Lab Invest 2001;81:919–928.

154 SCF MARKS MALIGNANT GERM CELLS

62. Shi SR, Key ME, Kalra KL. Antigen retrieval in formalinfixed, paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991;39:741–748. 63. Looijenga LH, Hersmus R, Gillis AJ, Pfundt R, Stoop HJ, van Gurp RJ, et al. Genomic and expression profiling of human spermatocytic seminomas: primary spermatocyte as tumorigenic precursor and DMRT1 as candidate chromosome 9 gene. Cancer Res 2006;66:290–302. 64. Zafarana G, Gillis AJ, van Gurp RJ, Olsson PG, Elstrodt F, Stoop H, et al. Coamplification of DAD-R, SOX5, and EKI1 in human testicular seminomas, with specific overexpression of DAD-R, correlates with reduced levels of apoptosis and earlier clinical manifestation. Cancer Res 2002;62:1822–1831. 65. McIntyre A, Summersgill B, Grygalewicz B, Gillis AJ, Stoop J, van Gurp RJ, et al. Amplification and overexpression of the KIT gene is associated with progression in the seminoma subtype of testicular germ cell tumors of adolescents and adults. Cancer Res 2005;65:8085–8089. 66. Satge D, Sommelet D, Geneix A, Nishi M, Malet P, Vekemans M. A tumor profile in Down syndrome. Am J Med Genet 1998;78:207–216. 67. He H, Jazdzewski K, Li W, Liyanarachchi S, Nagy R, Volinia S, et al. The role of microRNA genes in papillary thyroid carcinoma. Proc Natl Acad Sci USA 2005;102:19075–19080. 68. Sette C, Dolci S, Geremia R, Rossi P. The role of stem cell factor and of alternative c-kit gene products in the establishment, maintenance and function of germ cells. Int J Dev Biol 2000;44:599–608. 69. Ronnstrand L. Signal transduction via the stem cell factor receptor/c-Kit. Cell Mol Life Sci 2004;61:2535–2548. 70. Runyan C, Schaible K, Molyneaux K, Wang Z, Levin L, Wylie C. Steel factor controls midline cell death of primordial germ cells and is essential for their normal proliferation and migration. Development 2006;133:4861–4869. 71. Tu J, Fan L, Tao K, Zhu W, Li J, Lu G. Stem cell factor affects fate determination of human gonocytes in vitro. Reproduction 2007;134:757–765. 72. Looijenga LH, de Leeuw H, van Oorschot M, van Gurp RJ, Stoop H, Gillis AJ, et al. Stem cell factor receptor (c-KIT) codon 816 mutations predict development of bilateral testicular germ-cell tumors. Cancer Res 2003;63:7674–7678. 73. Biermann K, Goke F, Nettersheim D, Eckert D, Zhou H, Kahl P, et al. c-KIT is frequently mutated in bilateral germ cell tumours and down-regulated during progression from intratubular germ cell neoplasia to seminoma. J Pathol 2007;213:311–318. 74. Coffey J, Linger R, Pugh J, Dudakia D, Sokal M, Easton DF, et al. Somatic KIT mutations occur predominantly in seminoma germ cell tumors and are not predictive of bilateral disease: report of 220 tumors and review of literature. Genes Chromosomes Cancer 2008;47:34–42. 75. Hikono H, Zhou JH, Ohta M, Inumaru S, Momotani E, Sakurai M. Production of a monoclonal antibody that recognizes bovine stem cell factor (SCF) and its use in the detection and quantitation of native soluble bovine SCF in fetal bovine serum. J Interferon Cytokine Res 2002;22:231–235. 76. Shiohara M, Koike K, Kubo T, Amano Y, Takagi M, Muraoka K, et al. Possible role of stem cell factor as a serum factor: monoclonal anti-c-kit antibody abrogates interleukin-6-dependent colony growth in serum-containing culture. Exp Hematol 1993;21:907–912. 77. Rothschild G, Sottas CM, Kissel H, Agosti V, Manova K, Hardy MP, et al. A role for kit receptor signaling in Leydig cell steroidogenesis. Biol Reprod 2003;69:925–932. 78. Huyghe E, Soulie M, Escourrou G, Mieusset R, Plante P, Thonneau P. Conservative management of small testicular tumors relative to carcinoma in situ prevalence. J Urol 2005;173: 820–823. 79. Oosterhuis JW, Kersemaekers AM, Jacobsen GK, Timmer A, Steyerberg EW, Molier M, et al. Morphology of testicular parenchyma adjacent to germ cell tumours. An interim report. Apmis 2003;111:32–40; discussion 41–32.

155

A:IM>KØØ

<4:ABF4?

4AF&GBBC *<@8?F 8EG%B;?8 7>8E 4G;4E

Abstract Aims: Testis-sparing surgery might benefit quality of life, but can only be applied with histological examination for the presence of invasive germ cell tumour components, and the precursor carcinoma in situ (CIS). Currently, diagnosis is based on paraffin-embedded tissue, therefore a delay in further surgery is mainly unavoidable. The aim was to develop an intraoperative assessment technique using alkaline phosphatase as a marker.

Methods and results: A total of 4093 snap-frozen samples and matched paraffin- embedded tissue of 1500 patients were included. Besides standard haematoxylin and eosin (H&E) staining, the direct enzymatic alkaline phosphatase reactivity (dAP) test (duration 15 min) was applied on frozen sections, while H&E and immunohistochemistry for detection of OCT3 ⁄ 4, α-fetoprotein, human chorionic gonadotrophin (hCG) and cytokeratin was performed on the paraffin-embedded slides. Endothelial cells served as control for the dAP test. Positive staining was found in all CIS (n = 965), seminoma (n = 1035) and embryonal carcinoma (n = 584), either intratubular, microinvasive or invasive. Differentiated non- seminomas (n = 1238) showed variable staining. No staining was identified in spermatocytic seminomas (n = 5), testicular lymphomas (n = 42), testicular rhabdomyosarcomas (n = 7), Leydig cell tumours (n = 31), Sertoli-cell-only nodules (n = 4), (epi) dermoid cyst (n = 16), normal testicular parenchyma (n = 116), testicular torsion (n = 32) and inflammation of the epididymis (n = 19). The dAP test results matched H&E-stained parallel sections, as well paraffin-embedded tissue analysis, including immunohistochemistry.

Conclusions: The dAP test is an informative, reproducible and easy tool to diagnose CIS, (intratubular and microinvasive) seminoma and embryonal carcinoma on frozen tissue sections, being of great value in the context of sparing surgery.

158 ALKALINE PHOSPHATASE IN GCTS Introduction Testicular tumours are a heterogeneous group of lesions, with different origin and pathogenesis. These include Leydig cell tumours, Sertoli cell tumours, lymphomas, sarcomas and various types of tumours originating from the germ cell lineage.1 Based on clinical and pathobiological criteria, testicular germ cell tumours can be subdivided into three categories, i.e. teratomas and yolk sac tumours of neonates and infants (type I), seminomas and non-seminomas, referred to as TGCTs (type II), and spermatocytic seminomas (type III). In addition, there are benign teratomatous lesions with the histology of somatic tissue, including (epi)dermoid cysts. Although TGCTs are rare in the general population, with an incidence of between six and 11 per 100 000 in Caucasian males, they represent the most common cancer in men aged between 15 and 45 years and the incidence is still rising.2 High success rates in clinical management of TGCTs have been achieved using a combination of surgery, irradiation and chemotherapy.3 In fact, standard care for therapy for TGCTs is radical orchiectomy, which may be combined with irradiation and ⁄ or cisplatin-based chemotherapy, depending on the histology of the tumour as well as the disease stage. In the diagnostic work-up, the presence of elevated serum markers a- fetoprotein (AFP), ß-human chorionic gonadotrophin (ß-HCG) and ⁄ or lactate dehydrogenase (LDH) is informative. However, in present clinical practice a tumour in a testis is, in itself, enough reason to perform a radical orchiectomy, without a histological diagnosis of the testicular lesion. Recently, more attention has been given to the possibility of testis-sparing surgery in specific patients. This alternative procedure may be of interest for patients with a bilateral TGCT or an affected monotestis, in whom standard care would result in complete castration and lifelong dependence upon hormone substitution (testosterone replacement).4 To avoid these sequelae organ-sparing surgery might be an option, in which the tumour is enucleated and the normal appearing testis parenchyma is preserved.5,6 Other patients who may benefit from such an approach are those without elevated serum markers and a small, possibly benign tumour.7 However, a limitation is that no distinction can be made between a benign and a malignant lesion using the available imaging techniques, such as ultrasonography or magnetic resonance imaging. Therefore, by definition, exploratory surgery and frozen section analysis of tumour tissue is required. In fact, histological investigation should be performed of the lesion and its adjacent parenchyma.

Carcinoma in situ (CIS), also known as intratubular germ cell neoplasia unclassified (IGCNU)8 or testicular intratubular neoplasia (TIN), is the precursor of all TGCTs.9 Identification of CIS is sufficient for the diagnosis of a TGCT, which is by definition malignant. CIS is present in the majority of patients with an invasive

159 CHAPTER 11 TGCT, more frequently in patients with non-seminomas than with seminomas, which is probably related to the immune response.10 It is reported that 70% of patients with CIS will develop an invasive tumour within 7 years, which is extrapolated to 100% at 10 years.11 Of importance is the fact that CIS can be treated effectively by local low-dose irradiation with preservation of hormonal function of the testis in most cases.11 In daily practice, CIS is diagnosed using (formalin-, Stieve’s- or Bouin-) fixed, paraffin-embedded tissue, preferably supported by immunohistochemistry, for example (OCT)3 ⁄ 4,12–15 or c-KIT (CD117) and placental alkaline phosphatase (PLAP).8 The required time to diagnose CIS, based on paraffin-embedded tissue, possibly supported by immunohistochemistry, precludes its application as an intraoperative procedure. Therefore, if malignant cells are found in the adjacent parenchyma a second intervention, usually surgical, is required. Alternatively, frozen tissue diagnosis might be applied to determine the presence of CIS, as well as invasive tumour components, such as (intratubular and microinvasive) seminoma or embryonal carcinoma. The problem is that the morphology of frozen haematoxylin and eosin (H&E)-stained sections is often not sufficient to diagnose CIS reliably. This shortcoming of frozen section diagnosis could be compensated for with a specific marker that can be applied intra-operatively on frozen section. Alkaline phosphatase is such a marker.

Alkaline phosphatases are membrane-bound glycoproteins occurring in various forms and different tissues, which are encoded by different genes, i.e. tissue non- specific alkaline phosphatase (TNAP), tissue-specific alkaline phosphatase (TSAP), being placental-, intestine- and germ-cell alkaline phosphatase (PLAP and GCAP).16 No specific antibodies against all these various isozymes are currently available. Alkaline phosphatase is present in migrating primordial germ-cells (PGCs), and in gonocytes as well as in pre-spermatogonia in embryonal gonads.17 In fact, this enzymatic activity was used initially to identify mouse PGCs during normal embryogenesis.18,19 In normal testis parenchyma of adolescents and adults, no endogeneous activity of alkaline phosphatase is present, except in the endothelial cells of small blood vessels.20 It is known that PLAP and GCAP are expressed abundantly in seminoma and embryonal carcinoma.21,22

Because the enzyme is also present in CIS, it may be applied as a marker for the early diagnosis of TGCTs.17 The aim of this study was to develop the direct histochemical detection of alkaline phosphatase enzyme reactivity, which detects the activity of all enzymes, to diagnose CIS, (intratubular and microinvasive) seminoma and embryonal carcinoma on frozen tissue sections.

160 ALKALINE PHOSPHATASE IN GCTS Materials and methods The use of tissues for scientific purposes as included in this study was approved by an institutional review board (MEC 02.981 and CCR2041), and the samples were used according to the ‘Code for Proper Secondary Use’ (FMWV) (version 2002).

Primary tissue handling The material used was handled by routine procedures. Upon incision of the specimen a digital image was generated, and possible heterogeneity of the specimen identified. Representative parts of the tumour area(s) (possibly including different elements) and adjacent parenchyma were cut and fixed in 10% buffered formalin for maximum 24 h, and subsequently embedded in paraffin. Parallel areas of both the tumour and parenchyma were cut and snap- frozen in liquid nitrogen, according to special requirements. Small fresh biopsies need to be prepared as quickly as possible after surgical intervention. The biopsies were placed on a small piece of thin cork and covered with a small piece of wet filter paper, after which they were directly frozen in liquid nitrogen. Larger pieces of testis tissue were snap-frozen, without the cork support, again in liquid nitrogen. In addition, special care was taken in cutting the sections: under no circumstances could the frozen tissue be thawed, which would have compromised morphology and enzyme-histochemical staining and thus the interpretation of the slides.

The snap-frozen samples were investigated for histology using both the H&E staining and the dAP test (see below) on parallel slides, allowing optimal comparison. The diagnoses made based on the snap-frozen sections stained with either H&E or the dAP test were compared directly. The pathologist (W.O.) studied the H&E-stained sections(s) first and decided whether CIS and⁄or invasive tumour components were present. This was confirmed for CIS, (intratubular and micro-invasive) seminoma and embryonal carcinoma based on the dAP-stained slides. In addition, the findings on the frozen samples were compared to the diagnosis made on the formalin-fixed, paraffin-embedded tissues. Therefore, H&E staining was performed again, as well as immunohistochemistry using antibodies specifically detecting OCT3 ⁄ 4, AFP, hCG and cytokeratin (CAM5.2). All tumours were diagnosed according to the WHO classification.

161 CHAPTER 11 Direct alkaline phosphatase reactivity detection Frozen tissue sections of 5-lm thickness were air-dried (at room temperature for at least 30 s) and fixed for 10 s in 100% acetone. Subsequently, the slides were air-dried for 30 s and incubated for 5–10 min in 60 ml Tris HCl pH8.4 (0.2 m) with 20 mg Fast Red (Sigma Steinheim, Germany; F8764) and 20 mg Naphtol AS-MX phosphate salt (N5000; Sigma). After washing in running tapwater the slides were fixed for 30 s in 10% buffered formalin and counterstained with haematoxylin. Subsequently they were covered with an aqueous mounting medium (Imsolmount; Klinipath Duiven, The Netherlands; 4059) and inspected visually, after which they could be stored in the archive.

Results During a period of more than 15 years of collecting primary testicular tumours, in close collaboration with pathologists and urologists in the south-western part of the Netherlands, a large archive has been established. This tissue bank contains more than 1500 tumours from individual patients, from which frozen tissue as well as (matched) formalin-fixed, paraffin-embedded tissue is available. Representative slides of all frozen tissue samples were stained routinely using H&E and investigated for dAP; the findings showed no discrepancies. However, identification of CIS was more unequivocal using the dAP test. This was especially the case when only a relatively limited number of seminiferous tubules contained CIS cells, and when the affected tubules also retained spermatogenesis (see below). In addition, the findings were compared to the observations made on formalin-fixed paraffin-embedded tissues, which were also stained using H&E and investigated using immunohistochemistry (see Materials and methods section).

All cases were examined by an experienced pathologist with a special interest in TGCTs (J.W.O.). All tumours were described in a standardized, authorized pathology report in which various items were scored, including the presence of CIS, presence of the various invasive components and intratubular variants (seminoma and embryonal carcinoma). Again, no discrepancies were found between the various approaches. The results are summarized in Table 1, as well as the diagnosis made. Representative examples are shown in Figure 1.

Normal parenchyma and non-neoplastic conditions [normal parenchyma (Figure 1A), testicular torsion and chronic inflammation] showed no alkaline phosphatase reactivity, with the consistent exception of endothelial cells. The same pattern was found in all spermatocytic seminomas (Figure 1B) and non- TGCTs testicular lymphoma (Figure 1C), rhabdomyosarcoma, Leydig cell tumour (Figure 1D), Sertoli cell-only nodule, dermoid cysts and epidermoid cysts (Figure 1E). The endothelial cells, showing positive staining pre-dominantly of the 162 ALKALINE PHOSPHATASE IN GCTS smaller vessels, served as a positive internal control for this enzymatic staining. In contrast to the consistent negative finding as indicated above, all CIS (Figure 1F,G) and (intratubular and microinvasive) seminoma (Figure 1H,I) and embryonal carcinoma (Figure 1J,K) samples were positive. Interestingly, the CIS cells were not restricted to seminiferous tubules without spermatogenesis. The preinvasive tumour cells were located on the basal membrane in the niche beneath the tight junctions between the Sertoli cells. Some cases did not contain CIS, which was due to the absence of testicular parenchyma available for analysis, because of the overgrowth of the tumour. The same staining pattern as in CIS was seen in intratubular seminoma, although in this case the seminiferous tubule was completely filled with tumour cells, and Sertoli cells may have been completely absent.

Table 1. Results of direct alkaline phosphatase staining on frozen tissue sections of more than 4000 samples from 1500 patients. The number of cases investigated (=tested) and positive results (%) are indicated

Tested Histology positive

Benign lesions Normal parenchyma 116 0 Torsion 32 0 Chronic inflammation 19 0 Sertoli cell-only nodule 4 0

Non-TGCTs Spermatocytic seminoma 5 0 Lymphoma 42 0 Rhabdomyosarcoma 7 0 Leydig cell tumour 31 0 (Epi)Dermoid cyst 16 0

TGCTs and precursor Carcinoma in situ 965 100 Intratubular seminoma 244 100 Seminoma 791 100 Intratubular embryonal 71 100 carcinoma Embryonal carcinoma 513 100 Yolk sac tumour 402 * Choriocarcinoma 41 * Teratoma 795 *

*A heterogeneous staining pattern was identified, with both positive and negative regions, even within a single tumour. This observation does not negatively influence the power of the direct alkaline phosphatase method to detect carcinoma in situ, seminoma and embryonal carcinoma, both intra-tubular, microinvasive and invasive.

Upon invasive growth the staining pattern stayed the same. A different pattern was observed in embryonal carcinoma, in which not every single cell was positive. Instead, groups of tumour cells were defined by a polarized membranous staining for alkaline phosphatase. This pattern was similar for intratubular, microinvasive and frankly invasive embryonal carcinoma. In

163 CHAPTER 11 contrast, a more heterogeneously staining pattern of the enzymatic reactivity was identified in the various differentiated non-seminomatous components, such as yolk sac tumour (Figure 1L), choriocarcinoma (Figure 1M) and teratoma (Figure 1N). However, identification of the dAP test was not required in these cases, because diagnosis could be made based upon histological investigation (H&E-stained section) alone.

Gonadoblastoma, which is the CIS-counterpart of the dysgenetic gonad,23,24 also showed positive staining for alkaline phosphatase (Figure 1O, n = 7 independent cases). Recent examples illustrate the power of the method to guide intraoperative surgical decision-making. In two cases the suspicious testis (based on physical examination and ultrasound analysis) was not removed because staining for alkaline phosphatase ruled out the presence of a TGCT or CIS. In another case CIS and a small seminoma were demonstrated in a patient whose other testis had been removed earlier for a TGCT. During the operation it was decided to excise the seminoma and to spare the testicle to preserve its hormonal function. The remaining CIS was treated with local irradiation. The patient is under surveillance. An opportunity for testis-sparing surgery was missed in a 17-year-old boy. Microscopic examination of the removed testis demonstrated a small epidermoid cyst (11 mm in diameter), lacking an invasive component or CIS, thus constituting a benign lesion. In this case, testis-sparing surgery could have been performed if snap-frozen analysis had been carried out on both the tumour and the adjacent parenchyma to rule out the presence of CIS, (intratubular and microinvasive) seminoma and embryonal carcinoma.

Discussion Individualized treatment of cancer patients is becoming a major issue in current clinical practice. This is aimed at achieving optimal treatment success rates and at the same time reduction of treatment side effects to ensure quality of life, in both the short and long terms. Although intraoperative frozen section examination is applied commonly in surgical management for various malignancies, its role in clinical management of testicular masses is still in its infancy. However, in some cases organ-preserving surgery can be the preferred therapeutic intervention. These may include patients with a benign testicular lesion, as demonstrated by, for example, Heidenreich and coworkers in a series of 18 patients treated with enucleation of a benign tumour and analysis of frozen biopsies of the adjacent parenchyma.25

164 ALKALINE PHOSPHATASE IN GCTS

Figure 1. Examples of direct alkaline phosphatase staining on frozen tissue sections: (A) normal adult testis parenchyma; (B) spermatocytic seminoma; (C) lymphoma; (D) Leydig cell tumour; (E) epidermoid cyst, all showing only positive staining in endothelial cells; (F, G) carcinoma in situ (CIS) (parallel haematoxylin and eosin staining as insert, *position normal tubule); (H) intratubular seminoma; (I) invasive seminoma; (J) intratubular embryonal carcinoma; (K) invasive embryonal carcinoma, all showing a positive staining with a pattern as described in the text; (L) yolk sac tumour; (M) choriocarcinoma; (N) teratoma; (O) gonadoblastoma (showing a similar staining pattern to CIS).

165 CHAPTER 11 In addition, an adenomatoid tumour of the testis was treated by testis-sparing surgery after it was diagnosed as benign by intraoperative frozen section examination.26 Moreover, a segmental testicular infarction, simulating a TGCT, was treated by partial orchiectomy after a frozen section diagnosis.27

Currently, the standard treatment of a primary TGCT is radical orchiectomy, followed by additional treatment depending upon the stage of the disease and characteristics of the tumour, being either irradiation or chemotherapy.3 However, some patients might benefit from a more conservative surgical approach. This might prevent complete castration in the case of patients with a tumour in a solitary testis, a bilateral testicular tumour or a second testicular tumour after an initial unilateral orchiectomy. In fact, organ-preserving surgery is not yet standard care in these patients.28,29 Furthermore, sufficient endocrine function of the remaining testicular parenchyma should remain after the operation. If organ-preserving surgery is to be performed, detailed investigation of the adjacent testicular parenchyma is required to determine the presence of CIS, as well as (intratubular and microinvasive) seminoma and embryonal carcinoma. The other histological elements are easier to diagnose based on standard histological analysis. If CIS is documented histologically, adjuvant radiotherapy of the remaining testicular tissue is recommended strongly according to the management strategy for CIS in unilateral tumours.7 If invasive tumour components are identified, orchiectomy must be considered. Histological examination is performed predominantly on fixed tissue, which benefits from optimally preserved morphology. This is significantly reduced using snap-frozen tissue.

The procedure described in this paper has been investigated in a large prospective series of 1500 patients, in whom more than 4000 frozen and matched formalin-fixed paraffin-embedded samples were investigated. The dAP test allows fast and reliable identification of CIS (intratubular, microinvasive and invasive) seminoma and embryonal carcinoma on frozen tissue sections within 15 min of surgical removal. The absolute sensitivity and specificity of enzymatic alkaline phosphatase detection on frozen tissue makes identification of CIS, seminoma or embryonal carcinoma highly reproducible compared to standard snap-frozen tissue analysis based on H&E staining. The dAP test is able to detect even a single CIS cell in a seminiferous tubule with spermatogenesis, as well as a limited number of CIS containing seminiferous tubules in a background of mostly unaffected tubules. This method, which includes an internal positive control, i.e. endothelial cells, which were present in all investigated samples, was as informative as the immunohistochemical detection of OCT3⁄4 on paraffin- embedded tissue, based upon comparison of matched frozen and paraffin section analysis of all cases (data not shown). In fact, no discrepancies were

166 ALKALINE PHOSPHATASE IN GCTS found between the diagnoses made using the two methods. In addition, the test can be implemented easily in routine clinical histopathology.

The dAP test is a valuable diagnostic tool in testisconserving surgery, allowing direct evaluation of the presence of CIS, seminoma or embryonal carcinoma. With the availability of this intraoperative test the development and implementation of testis-sparing surgery is more feasible than hitherto. Even so, testissparing surgery should be performed at a centre with sufficient experience in the management of this rare clinical situation.30 However, some limitations must be borne in mind. Although CIS is found in most affected samples of testicular parenchyma, in the case of a type II GCT, it can show a heterogeneous pattern. This patchy distribution can result in a false negative finding upon biopsy analysis, which is not influenced by which detection technique is applied.16 In this context it is of relevance that two independent surgical biopsies are recommended for detection of CIS in a normal-sized testis, each with a diameter of at least 3 mm.31 However, although the risk will be small, the presence can never be absolutely excluded with a negative finding on biopsy material. Therefore, the proper size of the biopsy and subsequent handling and analysis are of great importance for optimal diagnosis.

References 1. Oosterhuis J, Looijenga L. Testicular germ-cell tumours in a broader perspective. Nat. Rev. Cancer 2005; 5; 210–222. 2. Bergstro¨m R, Adami H-O, Mohner M et al. Increase in testicular cancer incidence in six European countries: a birth cohort phenomenon. J. Natl Cancer Inst. 1996; 88; 727–733. 3. Horwich A, Shipley J, Huddart R. Testicular germ-cell cancer. Lancet 2006; 367; 754–765. 4. Ehrlich Y, Konichezky M, Yossepowitch O, Baniel J. Multifocality in testicular germ cell tumors. J. Urol. 2009; 181; 1114–1119. 5. Walsh C, Rushton HG. Diagnosis and management of teratomas and epidermoid cysts. Urol. Clin. North Am. 2000; 27; 509–518. 6. Heidenreich A, Weissbach L, Holtl W, Albers P et al. Organ sparing surgery for malignant germ cell tumor of the testis. J. Urol. 2001; 166; 2161–2165. 7. Schmoll HJ, Souchon R, Krege S et al. European consensus on diagnosis and treatment of germ cell cancer: a report of the European Germ Cell Cancer Consensus Group (EGCCCG). Ann. Oncol. 2004; 15; 1377–1299. 8. Woodward PJ, Heidenreich A, Looijenga LHJ, Oosterhuis JW, McLeod DG, Moller H Testicular germ cell tumors. In Eble JN, Sauter G, Epstein JI, Sesterhann IA eds. World Health Organization classification of tumours pathology and genetics of the urinary system and male genital organs. Lyon: IARC Press, 2004; 217–278. 9. Skakkebæk NE. Possible carcinoma-in-situ of the testis. Lancet1972; 2; 516–517. 10. Oosterhuis JW, Kersemaekers AM, Jacobsen GK et al. Morphology of testicular parenchyma adjacent to germ cell tumours; an interim report. APMIS 2003; 111; 32–42. 11. Rorth M, Rajpert-de Meyts E, Skakkebæk NE et al. Carcinoma in situ of the testis. Scand. J. Urol. 2000; 205; 166–186. 12. Looijenga LHJ, Stoop H, De Leeuw PJC et al. POU5F1 (OCT3 ⁄ 4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res. 2003; 63; 2244–2250.

167 CHAPTER 11

13 de Jong J, Stoop H, Dohle GR et al. Diagnostic value of OCT3 ⁄ 4 for pre-invasive and invasive testicular germ cell tumours. J. Pathol. 2005; 206; 242–249. 14. Van Casteren N, De Jong J, Stoop J et al. Evaluation of testicular biopsies for carcinoma in situ: immunohistochemistry is mandatory. Int. J.Androl. 2008; 32: 666–674, in press. 15. van Casteren NJ, Boellaard WP, Dohle GR et al. Heterogeneous distribution of ITGCNU in an adult testis: consequences for biopsy-based diagnosis. Int. J. Surg. Pathol. 2008; 16; 21–24. 16. Millan JL, Fishman WH. Biology of human alkaline phosphatases with special reference to cancer. Crit. Rev. Clin. Lab. Sci. 1995; 32; 1–39. 17. Hustin J, Collette J, Franchimont P. Immunohistochemical demonstration of placental alkaline phosphatase in various states of testicular development and in germ cell tumors. Int. J. Androl. 1987; 10; 29–35. 18. McGay G, Hertig AT, Adams EC, Danziger S. Histochemical observations on the germ cells of human embryos. Anat. Rec. 1953; 117; 201–219. 19. Chiquoine AD. The identification, origin, and migration of the primordial germ cells in the mouse embryo. Anat. Rec. 1954; 118; 135–146. 20. Roelofs H, Manes T, Millan JL, Oosterhuis JW, Looijenga LHJ. Heterogeneity in alkaline phosphatase isozyme expression in human testicular germ cell tumors. An enzyme- immunohistochemical and molecular analysis. J. Pathol. 1999; 189; 236–244. 21. Koshida K, Wahren BE. Enzymatic heterogeneity of seminomas. Clin. Chim. Acta 1990; 186; 255–264. 22. Koshida K, Uchibayashi T, Hisazumi H. Characterization of seminoma-derived placental-like alkaline phosphatase. Urol. Int. 1991; 47; 96–99. 23. Cools M, Drop SL, Wolffenbuttel KP, Oosterhuis JW, Looijenga LH. Germ cell tumors in the intersex gonad: old paths, new directions, moving frontiers. Endocr. Rev. 2006; 27; 468– 484. 24. Hersmus R, Wolffenbuttel KP, Drop SLS, Oosterhuis JW, Cools M, Looijenga LHJ. New insights into type II Germ Cell Tumor pathogenesis based on the studies of patients with various forms of Disorders of Sex Development (DSD). Mol. Cell. Endocrinol. 2008; 291; 1– 10. 25. Heidenreich A, Bonfig R, Derschum W, Von Vietsch H, Wilbert DM. A conservative approach to bilateral testicular germ cell tumors. J. Urol. 1995; 153; 10–13. 26. Mitsui Y, Ueda Y, Suzuki T et al. [A case of adenomatoid tumor of the testis treated by testis-sparing surgery: a case report]. Hinyokika Kiyo 2008; 54; 383–386. 27. Hidalgo J, Rodriguez A, Canalias J et al. Segmental testicular infarction vs testicular tumour: the usefulness of the excisional frozen biopsy. Arch. Esp. Urol. 2008; 61; 92–93. 28. Heidenreich A, Holtl W, Albrecht W, Pont J, Engelmann UH. Testis-preserving surgery in bilateral testicular germ cell tumours. Br. J. Urol. 1997; 79; 253–257. 29. Steiner H, Holtl L, Maneschg C et al. Frozen section analysisguided organ-sparing approach in testicular tumors: technique, feasibility, and long-term results. Urology 2003; 62; 508– 513. 30. Passman C, Urban D, Klemm K, Lockhart M, Kenney P, Kolettis P. Testicular lesions other than germ cell tumours: feasibility of testis-sparing surgery. BJU Int. 2009; 103; 488–491. 31. Dieckmann KP, Kulejewski M, Pichlmeier U, Loy V. Diagnosis of contralateral testicular intraepithelial neoplasia (TIN) in patients with testicular germ cell cancer: systematic two- site biopsies are more sensitive than a single random biopsy. Eur. Urol. 2007; 51; 175–183.

168

A:IM>KØÙ

?B54?!@8G;L?4G

BHEA4?B9#4G;B?B:LÒÒÛ ÔÓÓ ÔÔÒ ÒÚÛÚ

8A7E<>*8E@4AA 4AF&GBBC 78E %HH7I4AHEC "?8@@8ECB;? 8G4? ';8F84HG;BEF6BAGE<5HG87 8DH4??LGBG; CHAPTER 12

Abstract Differences in the global methylation pattern, ie hyper- as well as hypomethylation, are observed in cancers including germ cell tumours (GCTs). Related to their precursor cells, GCT methylation status differs according to histology. We investigated the methylation pattern of normal fetal, infantile, and adult germ cells (n = 103) and GCTs (n = 251) by immunohistochemical staining for 5- mcytidine. The global methylation pattern of male germ cells changes from hypomethylation to hypermethylation, whereas female germ cells remain unmethylated at all stages. Undifferentiated GCTs (seminomas, intratubular germ cell neoplasia unclassified, and gonadoblastomas) are hypomethylated, whereas more differentiated GCTs (teratomas, yolk sac tumours, and choriocarcinomas) show a higher degree of methylation. Embryonal carcinomas show an intermediate pattern. Resistance to cisplatin was assessed in the seminomatous cell line TCam-2 before and after demethylation using 5- azacytidine. Exposure to 5-azacytidine resulted in decreased resistance to cisplatin. Furthermore, after demethylation, the stem cell markers NANOG and POU5F1 (OCT3/4), as well as the germ cell-specific marker VASA, showed increased expression. Following treatment with 5-azacytidine, TCam-2 cells were analysed using a high-throughput methylation screen for changes in the methylation sites of 14000 genes. Among the genes revealing changes, interesting targets were identified: ie demethylation of KLF11, a putative tumour suppressor gene, and hypermethylation of CFLAR, a gene previously described in treatment resistance in GCTs.

170 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS Introduction DNA methylation is one of the major epigenetic modifications of the genome, occurring in both temporally and spatially defined patterns. In somatically differentiated cells, genomic methylation is in general stable and heritable [1]. It is involved in the regulation of gene expression, both of protein encoding and of non-encoding genes [2]. Both global demethylation and regional hypermethylation have been described in malignant cells [3]. Investigation of the global DNA methylation status on tissue sections using an antibody recognizing the presence of a methyl group on the 5’ carbon of cytidine (5-mC) has been reported [4]. This approach allows the examination of single cells within the histological context and at different stages of development and maturation. Here we report for the first time the global DNA methylation status of male and female fetal, infantile, and adult germ cells within their histological context in the gonads.

Human germ cell tumours (GCTs) originate from transformed primordial germ cells (PGCs)/gonocytes [5]. Although most cases show a unique sensitivity to cisplatin-based chemotherapy [6], GCTs are still a deadly disease in a minority of cases. The mechanisms underlying treatment failure are not well defined [7]. Methylation of CpG islands has been associated with drug sensitivity in several cell lines [8] and promoter methylation has been described in cisplatin resistance of testicular GCTs (TGCTs) [9]. In addition, we have demonstrated hypermethylation of the promoter of hMLH1 in cisplatin-resistant GCTs [10].

Using immunocytology, we analysed the global methylation pattern of cell lines and correlated it with levels of cisplatin resistance. In addition, the effects of chemically induced demethylation by 5-azacytidine on cisplatin sensitivity and the mRNA expression levels of a set of selected genes involved in differentiation were investigated. Furthermore, a set of 14 000 genes was analysed for changes in CpG island promoter methylation after 5-azacytidine exposure.

Materials and methods

Patients and samples Formalin-fixed, paraffin-embedded tissues of human normal fetal, infantile, and adult gonads including those from autopsies collected between 2000 and 2006 were analysed. Formalin-fixed, paraffin-embedded tissues of (T)GCT cases were collected between 1991 and 2001 in collaboration with urologists and pathologists in the south-western part of The Netherlands. In total, 251 cases of different histologies were available. All cases were reviewed by a pathologist with special experience in GCT pathology (JWO) and classified according to the WHO classification. Research on human tissue samples was performed according 171 CHAPTER 12 to the Code for Proper Secondary Use of Human Tissue in The Netherlands, as developed by the Dutch Federation of Medical Scientific Societies (FMWV), version 2002, and was approved by an institutional review board (MEC 02.981).

Immunohistochemistry Immunohistochemical detection of 5-mC was performed using a mouse monoclonal antibody (Eurogentec (Liège, Belgium), #BI-MECY-0100; diluted 1 : 500). Sections of the formalin-fixed, paraffin-embedded tissue were cut at 4 μm and treated, after deparaffinization and blocking of endogenous peroxidase, by pressure cooking (1200C, 1.2 bar) in Na citrate buffer (0.01 M, pH 6). Cytospin slides were formalin-fixed and subjected to the same antigen retrieval procedure as tissue sections. The slides were blocked for endogenous biotin and subsequently incubated at room temperature for 2 h with the primary antibody, for 30 min with a biotin-labelled secondary rabbit anti-mouse immunoglobulin, and again for 30 min with a biotinylated horseradish peroxidise– streptavidin complex (DAKO, Glostrup, Denmark and Vector, Burlingame, USA, respectively). 3,3’-Diaminobenzidine (DAB) was used as a chromogen. For immunohistochemistry of VASA, a rabbit polyclonal antibody (Abcam Cambridge, MA, USA, #ab13840; diluted 1 : 300) was used. No antigen retrieval was necessary for cytospins. Apart from the use of a biotinylated goat anti-rabbit secondary immunoglobulin, the same procedure as described above was applied.

Table 1. Global methylation pattern of different GCT subtypes and corresponding cell lines

Entity Global methylation pattern* n

Carcinoma in situ - 68 Gonadoblastoma - 11 Dysgerminoma - 8 Germinoma - 2 Seminomas Seminoma (control series) - 62 Resistant seminoma +++ 1 Seminoma cell line (TCam-2) +++ 1 Non-seminomas Embryonal carcinoma ++ 57 Yolk sac tumour +++ 51 Choriocarcinoma +++ 20 Mature teratoma +++ 50 Non-seminoma cell lines (NT2, 2102EP, NCCIT) +++ 3

* Results from 5-mC immunohistochemistry. -= no staining; ++= positive staining; +++= intense positive staining.

The results of 5-mC immunohistochemistry (only nuclear staining) on testicular samples were scored by defining the ratio of positive to negative cells per seminiferous tubule. Representative visual fields (200-fold magnification) were counted and the average number of negative cells was reported. On GCT

172 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS samples, three staining intensities were distinguished: no staining (-) when there was no positive nuclear signal in tumour cells (in the presence of a positive internal control); weak staining (++) in the case of an irregular pattern of staining in 20–80% of the tumour cell nuclei; and intense positive staining (+++) in the case of homogeneous positive tumour cell nuclei staining. All samples were scored independently by two investigators (HW and HS).

Western blot analysis

Protein isolation Pellets from the cell line TCam-2 were used to extract protein. Samples were washed in cold PBS and centrifuged at 00C for 5 min at 1500 rpm. Lysis buffer (1× PBS, 1% Nonidet P40, 0.5% sodium deoxycholate, 0.1% SDS) including protease inhibitors (complete mini-cocktail; Catalogue No 1 836 153, Roche Diagnostics, Mannheim, Germany) was added. Protein lysates were stored at -800C.

Western blot Samples of 10 μg protein were run on a 10% SDSpolyacrylamide gel, blotted on a PVDF membrane for 1.5 h, and incubated with the primary antibody for 2 h (anti- VASA rabbit polyclonal antibody, same as that used for immunohistochemistry; diluted 1 : 500) at room temperature. Binding of the primary antibody was visualized by using IRDye conjugated donkey anti-rabbit secondary antibody and the blots were scanned on the Odyssey infrared imaging system (LI COR Biosciences, NE, USA). As loading control, anti-ß-actin was used (Abcam, #ab8226; diluted 1 : 50 000).

MTT vitality assay on cisplatin exposure with and without 5-azacytidine treatment The effect of cisplatin on TCam-2 cell number was assessed using the MTT (3- [4,5-dimethylthiazol-2-yl]- 2,5-diphenyltetrazolium bromide) colorimetric assay as described previously [11]. TCam-2 cells were plated in 96-well plates at a density of 2 × 103 cells/100 μl medium per well in eight-fold replicates and allowed to adhere. After 24 h, TCam-2 cells were exposed to medium either with or without addition of 5-azacytidine (10 μM final concentration). After another 24 h pre-incubation, cisplatin was added in different concentrations and after 72 h, MTT (Sigma Chemical Co, St Louis, MO, USA) was added at a final concentration of 0.5 mg/ml. Cells were further incubated for 2 h. Plates were centrifuged for 5 min at 3000 rpm and the medium was carefully aspirated. Solubilization of formazan crystals was achieved by adding SDS (Sigma) and gentle shaking of the plates for 10 min. Finally, absorbance was read at 570 nm and results were expressed as the percentage of treated (cisplatin) versus untreated cells.

173 CHAPTER 12 mRNA expression profiling

Quantitative RT-PCR Total RNA was extracted from lysates of TCam-2 and TCam-2 treated with 5- azacytidine (24 h, 10 μM) using an RNAqueous-4PCR kit (Ambion/Applied Bio- systems, Nieuwerkerk a/d IJssel, The Netherlands) according to the manufacturer’s instructions. Samples were treated with DNase-I and checked for residual DNA contamination by PCR using a primer set for SOX2, which is a single exon gene. cDNA was generated as previously described [12]. Only samples found to be DNA-negative were included for further investigation. Quantitative PCR was performed using the 7900HT Fast Real-Time PCR System (Applied Biosystems). A specific set of primers was used to detect the expression of a set of genes known to be related to various differentiation lineages (see Supporting infor-mation, Supplementary Table 1). Quantitative values were obtained from the Cts as follows: target mRNA was quantified relative to HPRT (target mRNA value = 2(mean Ct HPRT-Ct target)) [12].

High-throughput methylation profiling The seminomatous cell line TCam-2 was analysed by high-throughput methylation profiling before and 22 passages after chemical demethylation with 5-azacytidine (one time exposure for 24 h; 10 μM final concentration). DNA bisulphite conversion was per-formed using the Zymo EZ DNA Methylation Kit (Zymo Research, Orange, CA, USA) according to the manufacturer’s protocol with the modifications described in the Infinium Assay Methylation Protocol Guide (Illumina Inc, San Diego, CA, USA). Sub-sequent analysis steps were performed according to the Infinium II Assay Lab Setup and Procedures and the Infinium Assay Methylation (http://www.illumina.com/technology/infinium_ methylation assay.ilmn; accessed August 2009) Protocol Guide (http://www.Illumine.com/ products/infinium_humanmethylation27_beadchip_kits.ilmn#documentation; accessed August 2009) (Illumina Inc). The processed DNA samples were hybridized to the HumanMethylation27 DNA Analysis BeadChip (http://www.illumina.com/products/infinium_humanmethylation27_beadchip_ leits.ilmn; accessed August 2009) (Illumina Inc), which allows assay at 27578 CpG sites selected from more than 14000 genes. Differential methylation analysis (DMA) was performed using the BeadStudio software (default settings; version 3.1.3.0, Illumina Inc). The global DNA methylation data showed a bimodal distribution in which beta values below 0.25 define the unmethylated CpGs and beta values above 0.75 the methylated CpGs. Thus, beta values below 0.25 and above 0.75 were selected as threshold values to define unmethylated and methylated CpG loci for further analysis. The thresholds for hyper- and hypomethylation have been established previously [13–15]. CpG loci were classified as differentially methylated if they showed a significant difference (< 0.001,

174 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS corresponding to a DiffScore of less than -30 or more than 30) in their DNA methylation values of at least 0.3 (delta-beta >0.3) between the samples analysed. The gene detection rate (p < 0.01) was high: 27552 of 27568 (99.94%; TCam-2 untreated) and 27548 of 27568 (99.93%; TCam-2 5-azacytidine treated), respectively, demonstrating a high quality of the sample DNA and the experimental performance. The reproducibility and accuracy of the HumanMethylation27 Bead Array technology has been demonstrated in several analyses published before using different techniques: eg bisulphite sequencing, bisulphite pyrosequencing, methylation-specific PCR or GoldenGate analysis [13,16–18].

Figure 1. Methylation status of male and female gonads during pre- and post-natal development. (A) Representative examples of 5-mC immunostaining of fetal and mature female gonadal tissue (23 weeks gestational age: left image; 3 years: right image). (B) Methylation pattern in testicular tissue. Upper panel: schematic representation of absolute numbers of cells per seminiferous tubule showing no immunohistochemical staining for 5-mC at different weeks of gestational age (median time of birth: 40 ± 2); lower panel: representative examples of 5-mC immunostaining at the indicated time points (arrows: germ cells).

175 CHAPTER 12 Results

5-mC detection by immunostaining in fetal and adult gonadal tissue The global methylation pattern of normal fetal germ cells during maturation from fetal to adult stages was assessed by 5-mC immunostaining. Both fetal, infantile testes (n = 45) and ovaries (n = 19) starting from week 15 gestational age to post-pubertal age were analysed. In female gonads, germ cells, ie oogonia, oocytes, and follicles at different stages of maturation, are always unmethylated, both prenatally and after birth (see Figure 1A for representative illustrations). A more heterogeneous pattern was found in testes. Germ cells in fetal testes at early stages (starting from 15 weeks gestational age onwards) are predominantly hypomethylated (see Figure 1B for the scoring results and representative illustrations). With time, the number of demethylated germ cells declines and after birth, almost all germ cells are methylated. Interestingly, around the time of birth (week 40 ± 2), a short rise in the number of germ cells showing global demethylation is observed.

5-mC detection by immunostaining in germ cell tumours A distinct methylation pattern is found in undifferentiated and more differentiated TGCTs. Immunohistochemical findings are presented in Table 1 and Figure 2. The known precursor lesions intratubular germ cell neoplasia unclassified (IGCNU) of the testis and gonadoblastoma of the dysgenetic gonad, which morphologically resembles seminomatous cells, are both demethylated (Figures 2A and 2B). In addition, all seminomas are demethylated (Figure 2C). This pattern can also be seen in histologically similar tumours originating in the brain (germinomas) and the ovary (dysgerminomas), all of which are demethylated (data not shown). While embryonal carcinomas show an intermediate pattern (Figure 2E), the more differentiated non-seminomas, ie yolk sac tumours, choriocarcinomas and teratomas, are consistently hypermethylated (Figure 2F). Overall, a clear correlation between differentiation and methylation can be described: more differentiated tumour histologies show a higher degree of methylation. In contrast to a large series of unselected seminomas all showing demethylation, a single case of a relapsed seminoma after cisplatin treatment from our archive was not demethylated, but showed an atypical pattern of hyper-methylation (see Figure 2D). The only available cell line with seminomatous characteristics (TCam-2) as well as three embryonal carcinoma cell lines showed a high level of methylation (Figures 2G and 2H).

176 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS

Figure 2. Representative examples of 5-mC immunostaining of the different GCT entities and corresponding cell lines: (A) intratubular germ cell neoplasia unclassified (IGCNU); (B) gonadoblastoma. Note that the positive cells are supportive cells;(C) seminoma;(D) seminoma showing relapse after cisplatin-based therapy. Note that in both cases, the most intensive staining is observed in infiltrating lymphocytes, which can be used as an internal staining control;(E) embryonal carcinoma;(F) mature teratoma;(G) seminoma cell line TCam-2;(H) embryonal carcinoma cell line NT2.

Figure 3. Cisplatin resistance levels of the seminoma cell line TCam-2. Vitality curves of untreated (green) and 5-azacytidine pretreated (red: 24 h; brown: 24 h + five passages) TCam-2 cells under cisplatin exposure for 72 h. Immunohistological images are shown as insets of matching cytospins stained for 5-mC displaying the respective methylation state. All curves represent the results of eight- fold analyses (standard deviation indicated in bars). The assays were repeated in three independent experiments, yielding comparable results.

177 CHAPTER 12 Cisplatin resistance level of the seminoma cell line TCam-2 before and after 5-azacytidine treatment To investigate the influence of the methylation status on the resistance level of the seminoma cell line TCam-2, vitality assays were performed under cis-platin incubation with or without pretreatment with 5-azacytidine. TCam-2 showed an

IC50 concentration of cisplatin of 12 μM after 72 h of treatment (see Figure 3). Upon pretreatment with 5-azacytidine for 24 h, a significant decrease in resistance to cisplatin (IC50 of 8 μM) was observed. This effect remained stable in TCam-2 cells after further passaging with-out further 5-azacytidine treatment: the IC50 cisplatin concentration decreased from 12 to 7 μM after five passages in culture. Cytospins were carried out to confirm the efficiency of the demethylation procedure; they demonstrated that the cells were predominantly demethylated and remained so up to the eighth passage after a single 5 azacytidine exposure (see also Figure 3, insets).

Gene expression changes in TCam-2 after 5-azacytidine treatment The effect of 5-azacytidine treatment was investigated by quantitative RT-PCR analysing the transcript levels of a set of 29 genes related to differentiation. The absolute values (normalized to an internal control using the housekeeping gene HPRT of the untreated cell line) are shown in Figure 4A and are in line with data from previous genome-wide expression profiling [19]. The relative changes after 5-azacytidine treatment in comparison to the untreated cell line are shown in Figure 4B. The pluripotency markers NANOG and POU5F1 (OCT3/4 ), already strongly expressed in untreated TCam-2 cells, further increased in expression. In addition, the expression of several markers associated with differentiation, such as TEAD1 (late trophectoderm), BMPR1A (cardiac differentiation), and SMAD4 (signalling), increased after 5-azacytidine treatment. The most distinguishing marker between untreated and treated TCam-2 cells is the germ cell-specific marker VASA, showing a 16-fold increase in expression in demethylated cells. No changes in histomorphology or cell growth characteristics were observed in vitro (data not shown).

VASA protein expression changes after 5-azacytidine treatment To verify the observed changes in gene expression, western blot analyses were performed for VASA (see Figure 4C). In line with the mRNA expression levels, the amount of VASA protein increased in lysates from demethylated TCam-2 cells, albeit to a lesser extent than on the mRNA level. A 17% gain of signal in digital image quantification was detected compared with the original TCam-2 (Figure 4C, left panel). This was also confirmed by immunohistochemical staining of cytospins for VASA, which showed stronger expression in cells after exposure to 5-azacytidine (Figure 4C, right panel).

178 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS Changes in the methylation of 27 578 CpG sites in the seminoma cell line TCam-2 after 5-azacytidine treatment Upon continuous passaging after 5-azacytidine treatment, initially demethylated TCam-2 cells showed, from the tenth passage onwards, regeneration of their level of global methylation (immunohistochemical detection; data not shown). Interestingly, however, they remained more sensitive to cisplatin than untreated cells. This phenomenon was further analysed using a high-throughput methylation screen of 27 578 CpG sites, representing a subset of 14 000 genes. TCam-2 cells at passage 22 after single exposure to 5-azacytidine treatment showed significant hypomethylation in 54 CpG sites (23.5%), whereas 176 loci (76.5%) showed hypermethylation. Figure 5 shows the change in methylation of all analysed CpG sites, high-lighting the 15 most hyper- and de-methylated sites. Remarkably, the tumour suppressor gene KLF11 was the most highly demethylated CpG site, whereas the top three hypermethylated CpG sites include two proto-oncogenes, namely CFLAR and ERBB2.

Discussion In this study, we demonstrate that the global methylation state of fetal human germ cells shows a specific pattern of temporal regulation and is different in female and male gonads. During testicular development, the pattern changes from a generally hypomethylated status in fetal gonads to a hypermethylated status in more mature germ cells. In contrast, female germ cells remain unmethylated during all developmental stages investigated. Previously, it has been reported that PGCs, the cell of origin of both female and male mature germ cells, are completely unmethylated after undergoing a methylation erasure [2,20]. As all type II TGCTs are believed to originate from a common precursor, a transformed PGC/gonocyte, which has not been re-programmed by de novo methylation, the global methylation status of particular TGCT subtypes is supposed to differ according to the time point of their developmental arrest [21– 23]. It has been reported that the global methylation status of non-seminomas is different from both seminomas and IGCNU [24]. In line with these findings, our investigation shows that the precursor lesions IGCNU and gonadoblastomas and (T)GCTs that resemble PGCs/gonocytes, ie semi-nomas, germinomas, and dysgerminomas, are largely unmethylated. Non-seminomatous GCTs have a global methylation status according to their degree of differ-entiation. Embryonal carcinoma, the undifferentiated non-seminoma component, shows intermediate levels of methylation, while choriocarcinomas, yolk sac tumours, and teratomas are highly methylated.

179 CHAPTER 12

Figure 4. qRT-PCR results from a set of 29 genes investigated in untreated TCam-2 (yellow), TCam-2 after demethylation by single 5-azacytidine treatment (24 h) (red), and TCam-2 eight passages after demethylation by 24 h treatment with 5-azacytidine (green). (A) Absolute values are normalized to HPRT. (B) Results are depicted as changes in gene expression after demethylation, expressed as a ratio per gene compared with untreated TCam-2. (C) Immunoblots detecting VASA protein in TCam-2 cell lysates, untreated and after demethylation, passage 17 after 5-azacytidine exposure. Corresponding cytospins showing VASA staining are shown. VASA signal is detected by DAB (brown nuclear and cytoplasmic signal).

180 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS

Figure 5. Results from methylation profiling of 27 578 CpG sites in TCam-2 cells. Changes in methylation after 22 passages after demethylation by 24 h treatment with 5-azacytidine are depicted. The pie diagram shows changes in the percentages of significant deviating CpG sites (see text for further explanation). The 15 CpG sites showing the highest degree of hyper- and demethylation are highlighted.

Another distinguishing factor between seminomas and non-seminomas with evident clinical implication is their responsiveness to chemotherapy, mainly cisplatin. Seminomatous GCTs usually show an excellent response to cisplatin, whereas the response of non-seminomatous GCTs to chemotherapy can differ according to histology: highly differentiated mature teratomas show no response to cisplatin [25,26]. Furthermore, in non-seminomatous GCTs, cisplatin resistance is a clinical problem, rendering the tumour a deadly disease in a minority of cases

181 CHAPTER 12 [27–29]. Our observations suggest an association between differentiation, global DNA methylation status, and response to chemotherapy. Regarding global DNA methylation in seminomas, a hypermethylated genome was found in only one case showing relapse after cisplatin-based chemotherapy, whereas no other seminoma out of 62 cases from our tumour archive showed hypermethylated DNA. Clearly, this finding warrants further investigation of the methylation status of relapsed seminomas in a larger series. Interestingly, the only available cell line with seminoma characteristics, TCam-2, which compared to nonseminoma cell lines shows surprisingly high resistance to cisplatin (IC5012 μM versus 0.2–5 μM, according to our own findings [30,31]), shows a high degree of DNA methylation. In our view, hypermethylation of TCam-2 could be a cause of the high cisplatin resistance in this cell line, which shares many similarities with seminomas [19] with the exception of the high level of methylation. Therefore TCam-2 can be regarded as an in vitro model for analysing mechanisms of resistance in seminomas. The impact on resistance of non-seminomatous components needs to be investigated.

To investigate the role of DNA methylation in cisplatin resistance, we analysed the effects of the demethylating agent 5-azacytidine on treatment resistance. Chemical global demethylation of TCam-2 cells led to a decrease in cisplatin resistance. Most recently, a study reported that treating embryonal carcinoma cell lines with 5-azacytidine induces cisplatin sensitivity [32]. Our findings are in line with this observation and describe the same effect in a seminoma cell line, making a general susceptibility of all GCTs to treatment with 5-azacytidine very likely.

To reveal differences in gene expression that could be associated with the observed change in cisplatin resistance, a panel of 29 markers of differentiation was analysed by quantitative RT-PCR in TCam-2 cells treated with or without 5- azacytidine. Upon demethylation, the expression of the pluripotency markers NANOG and OCT3/4 increased, reflecting an association between hypomethylation and an undifferentiated state. However, no gross morphological changes in TCam-2 cells were observed in vitro. The most distinguishing marker between methylated and unmethylated TCam-2 cells is the germ cell-specific marker VASA. Whether expression of VASA has a direct impact on cisplatin resistance is unclear and warrants further investigations. In general, it seems likely that the observed change in cisplatin resistance levels after global demethylation involves the regulation of many genes and is therefore multi-factorial.

To further analyse potential target genes involved in cisplatin resistance, we performed an analysis of more than 14 000 genes by a high-throughput

182 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS methylation screen. In total, 27 578 CpG island sites of TCam-2 cells 22 passages after a single course of 5-azacytidine treatment were analysed and compared with CpG sites of untreated cells. Surprisingly, the vast majority of analysed CpG sites which show a significant change in the methylation status became hypermethylated with time, compared with untreated cells, whereas only a few remained demethylated (76.5% versus 23.5%, respectively). Among the genes showing the most significant changes in methylation, three genes seem particularly noteworthy: the most demethylated CpG site maps to KLF11, a putative tumour suppressor gene, whereas among the top three hypermethylated CpG sites, two proto-oncogenes were identified: CFLAR (c-FLIP ) and ERBB2. Functionally, it remains to be elucidated in which way these genes are involved in cisplatin resistance in GCTs. Interestingly, CFLAR has been described among the top ten genes predicting overall survival in a recent analysis of tumour samples from GCT patients [33]. The c-FLIP protein is a direct inhibitor of active caspase 8 formation and therefore of executing apoptosis. A regulatory role of caspase activation has been reported for cisplatin resistance in TGCTs [34]. Furthermore, it has been shown in a cervical cancer cell line in vitro as well as in colorectal cancer in vivo that c-FLIP triggers response to cisplatin-based chemotherapy [35,36].

Conclusions Germ cells at different stages of maturation as well as (T)GCTs of different histologies show marked differences in their global DNA methylation pattern. More differentiated cells show a higher degree of DNA methylation. The more undifferentiated GCTs such as seminomas and dysgerminomas are mainly demethylated, whereas non-seminomas show more methylated DNA. Interestingly, we found a single case of a seminoma after relapse from cisplatin- based chemotherapy that showed an unusual hypermethylation of the DNA in contrast to a large control series of seminomas. In vitro, chemical demethylation with 5-azacytidine decreased the intrinsically high resistance to cisplatin in TCam-2, a seminoma cell line. Further in vitro and in vivo experiments are warranted to explore the promising potential of demethylating agents such as 5- azacytidine in the treatment of cisplatin-resistant (T)GCTs.

Acknowledgment We thank the pathologists and urologists in the south-western part of The Netherlands for their support in collecting the tumour samples. HW was supported by an ESMO grant and FH was supported by Werner Otto Stiftung.

183 CHAPTER 12 References 1. Reik W, Dean W, Walter J. Epigenetic reprogramming in mammalian development. Science (NY) 2001; 293: 1089–1093. 2. Biermann K, Steger K. Epigenetics in male germ cells. J Androl 2007; 28: 466–480. 3. Esteller M. Cancer epigenomics: DNA methylomes and histone-modification maps. Nature Rev Genet 2007; 8: 286–298. 4. Piyathilake CJ, Johanning GL, Frost AR, et al . Immunohistochemical evaluation of global DNA methylation: comparison with in vitro radiolabeled methyl incorporation assay. Biotech His-tochem 2000; 75: 251–258. 5. Oosterhuis J, Looijenga L. Testicular germ-cell tumours in a broader perspective. Nature Rev Cancer 2005; 5: 210–222. 6. Einhorn LH. Curing metastatic testicular cancer. Proc Natl Acad Sci U S A 2002; 99: 4592– 4595. 7. Mayer F, Honecker F, Looijenga LHJ, et al . Towards under-standing the biological basis of the response to cisplatin-based chemotherapy in germ cell tumors. Ann Oncol 2003; 9: 825–832. 8. Shen L, Kondo Y, Ahmed S, et al . Drug sensitivity prediction by CpG island methylation profile in the NCI-60 cancer cell line panel. Cancer Res 2007; 67: 11335–11343. 9. Koul S, McKiernan JM, Narayan G, et al . Role of promoter hypermethylation in cisplatin treatment response of male germ cell tumors. Mol Cancer 2004; 3: 3–16. 10. Honecker F, Wermann H, Mayer F, et al . Microsatellite instability, mismatch repair deficiency, and BRAF mutation in treatment-resistant germ cell tumors. J Clin Oncol 2009; 27: 2129–2136. 11. Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983; 65: 55–63. 12. Looijenga LHJ, Hersmus R, Gillis A, et al . Genomic and expression profiling of human spermatocytic seminomas; primary sper-matocyte as tumorigenic precursor and DMRT1 as candidate chro-mosome 9-gene. Cancer Res 2006; 66: 290–302. 13. Martin-Subero JI, Kreuz M, Bibikova M, et al . New insights into the biology and origin of mature aggressive B-cell lymphomas by combined epigenomic, genomic, and transcriptional profiling. Blood 2009; 113: 2488–2497. 14. Martin-Subero JI, Ammerpohl O, Bibikova M, et al . A comprehensive microarray-based DNA methylation study of 367 hematological neoplasms. PLoS One 2009; 4: e6986. 15. Richter J, Ammerpohl O, Martin-Subero JI, et al . Array-based DNA methylation profiling of primary lymphomas of the central nervous system. BMC Cancer 2009; 9: 455. 16. Ammerpohl O, Martin-Subero JI, Richter J, et al . Hunting for the 5th base: techniques for analyzing DNA-methylation. Biochim Biophys Acta 2009; 1790: 847–862. 17. Kanber D, Berulava T, Ammerpohl O, et al . The human retinoblastoma gene is imprinted. PLoS Genet 2009; 5: e1000790. 18. Ammerpohl O, Martin-Subero JI, Richter J, et al . Genome-wide detection of DNA methylation changes in lymphomas using the Infinium HumanMethylation27 BeadChip. iCommunity Newslet-ter —Illumina customer journal. 2008; May 1. 19. de Jong J, Stoop H, Gillis AJ, et al . Further characterization of the first seminoma cell line TCam-2. Genes Chromosomes Cancer 2008; 47: 185–196. 20. Lind GE, Skotheim RI, Lothe RA. The epigenome of testicular germ cell tumors. APMIS 2007; 115: 1147–1160. 21. Okamoto K, Kawakami T. Epigenetic profile of testicular germ cell tumours. Int J Androl 2007; 30: 385–392; discussion 92. 22. Reuter VE. Origins and molecular biology of testicular germ cell tumors. Mod Pathol 2005; 18 (Suppl 2): S51–S60.

184 DNA METHYLATION, DIFFERENTIATION AND RESISTANCE IN GCTS

23. Smiraglia DJ, Szymanska J, Kraggerud SM, et al . Distinct epigenetic phenotypes in seminomatous and nonseminomatous testicular germ cell tumors. Oncogene 2002; 21: 3909–3916. 24. Netto GJ, Nakai Y, Nakayama M, et al . Global DNA hypomethylation in intratubular germ cell neoplasia and seminoma, but not in nonseminomatous male germ cell tumors. Mod Pathol 2008; 21: 1337–1344. 25. Krege S, Beyer J, Souchon R, et al . European Consensus Conference on Diagnosis and Treatment of Germ Cell Cancer: a report of the Second Meeting of the European Germ Cell Cancer Consensus group (EGCCCG): Part I. Eur Urol 2008; 53: 478–496. 26. Krege S, Beyer J, Souchon R, et al . European Consensus Conference on Diagnosis and Treatment of Germ Cell Cancer: a report of the Second Meeting of the European Germ Cell Cancer Consensus Group (EGCCCG): Part II. Eur Urol 2008; 53: 497–513. 27. Kollmannsberger C, Honecker F, Bokemeyer C. Pharmacotherapy of relapsed metastatic testicular cancer. Expert Opin Pharmacother 2008; 9: 2259–2272. 28. Hartmann JT, Einhorn L, Nichols CR, et al . Second-line chemotherapy in patients with relapsed extragonadal non-seminomatous germ cell tumors: results of an international multicenter analysis. J Clin Oncol 2001; 19: 1641–1648. 29. Hartmann JT, Nichols CR, Droz JP, et al . Prognostic variables for response and outcome in patients with extragonadal germ-cell tumors. Ann Oncol 2002; 13: 1017–1028. 30. Koch S, Mayer F, Honecker F, et al . Efficacy of cytotoxic agents used in the treatment of testicular germ cell tumours under normoxic and hypoxic conditions in vitro. Br J Cancer 2003; 89: 2133–2139. 31. Glaesener S, Honecker F, Veltman IM, et al . Comparative proteome, transcriptome, and genome analysis of a gonadal and an extragonadal germ cell tumor cell line. J Proteome Res 2008; 7: 3890–3899. 32. Beyrouthy MJ, Garner KM, Hever MP, et al . High DNA methyl-transferase 3B expression mediates 5-aza-deoxycytidine hyper-sensitivity in testicular germ cell tumors. Cancer Res 2009; 69: 9360–9366. 33. Korkola JE, Houldsworth J, Feldman DR, et al . Identification and validation of a gene expression signature that predicts outcome in adult men with germ cell tumors. J Clin Oncol. 2009; 27: 5240–5247. 34. Mueller T, Voigt W, Simon H, et al . Failure of activation of caspase-9 induces a higher threshold for apoptosis and cisplatin resistance in testicular cancer. Cancer Res 2003; 63: 513–521. 35. Luo A, Wang W, Sima N, et al . Short hairpin RNA targeting c-FLIP sensitizes human cervical adenocarcinoma HeLa cells to chemotherapy and radiotherapy. Cancer Lett 2008; 271: 323–332. 36. Longley DB, Wilson TR, McEwan M, et al . c-FLIP inhibits chemotherapy-induced colorectal cancer cell death. Oncogene 2006; 25: 838–848.

SUPPORTING INFORMATION ON THE INTERNET The following supporting information may be found in the online version of this article. Table S1. Details of quantitative RT-PCR analysis of 29 markers related to differentiation.

185

A:IM>KØÚ

8A8E4?

CHAPTER 13

13.1 Introduction Histologically type II (T)GCTs mimick embryogenesis to a certain extent, although in an uncontrolled manner. The parallels between normal embryogenesis and tumorigenesis strengthen the need to understand mechanisms involved in early development, especially related to the different developmental stages of the gonads, including the germ cells. The studies described in this thesis deal with expression profiles of proteins during normal human testicular development and the translation of the observations into clinical practice, in particular for the diagnostic workup of (T)GCTs. The obtained results will be discussed in the context of optimal early diagnosis, as well as understanding the pathogenesis of this disease.

13.2 Normal testicular development and protein expression profile Germ cells during normal development show a number of defined steps of maturation. They originate as PGCs, and after completing the process of spermatogenesis, induced after puberty, spermatozoa are formed, consisting of highly specialized cellular compositions and a haploid set of chromosomes, suitable for fertilization. However, already during embryonal and neonatal development, various maturation stages of male germ cells are defined. These are of specific relevance in the context of (T)GCTs, and were therefore the main topic of the investigations performed and described here. After migration of PGCs from the yolk sac to the genital ridge, the germ cells are referred to as gonocytes, which are associated with the supportive cells, being the Sertoli precursors cells. At this moment of development, the germ cells are positive for a number of markers, including PLAP, VASA, c-KIT, TSPY, and OCT3/4 1. Interestingly, under physiological conditions, TSPY is limited in expression to germ cells in a testicular micro-environment. Co-expression of these markers can identify various subpopulations of germ cells, supporting a stepwise maturation process 1-5. However, the impact of the different subpopulations is unknown so far. Subtle differences in the identified subpopulations of germ cells are reported regarding their individual marker staining profiles 6-9. Again, the value remains to be determined.

During subsequent development, the germ cells, which are initially located in the central position of the developing seminiferous tubule, move to a location at the basal lamina. These will eventually home to the niche under the tight junctions of the Sertoli cells, at which stage they are referred to as pre-spermatogonia. In association with this transition, the germ cells show a reduced protein expression of PLAP, c-KIT, and OCT3/4, while VASA and TSPY remain positive. In fact, PLAP

188 GENERAL DISCUSSION and OCT3/4 disappear completely, while c-KIT remains detectable depending on the method of fixation and detection, even in an adult (post-pubertal) testis (unpublished observations). A similar profile as the embryonic germ cell markers is found for the proliferative activity, using Ki-67 as marker. Interestingly, both VASA and TSPY are informative as markers for the presence of germ cells, independent of the developmental age of the testis. It is of importance to note that VASA is detected at multiple maturation stages, while TSPY is restricted to the pre-spermatogonial germ cells. . This is schematically Illustrated in Figure 1A.

In conclusion, after birth, the embryonic germ cell markers can be sporadically found, although they are absent after 6-12 month post partum, associated with a decline in proliferative activity.

Figure 1. Protein expression in normal and aberrant germ cell development. (A) Expression of the different diagnostic markers alkaline phosphatase (AP), OCT3/4, c-KIT, VASA, SOX17 and TSPY during the normal development of the male gonad from fertilization till puberty. No detectable expression of SCF. (B) The expression of the different diagnostic markers during aberrant germ cell development, leading to CIS, is shown. There is an ongoing expression of the embryonal diagnostic markers after birth in CIS cells. This is in contrast with the normal expression in which about 6-12 months after birth all the proteins related to the embryonal stage are disappeared with an exception for VASA and TSPY. A significant difference is the expression of SCF only detectable around the niche of the (pre)CIS cells.

189 CHAPTER 13

Besides based on protein expression profiling also regarding epigenetic (DNA) modification, different stages of germ cell development can be identified. Using immunohistochemical detection of 5-MC, the PGC/gonocyte stage is overall demethylated, while during transition towards pre-spermatogonia the level of methylation is increased, which is retained in later stages. The specific timing of down-regulation of the embryonic markers in germ cells in association to the methylation status is not investigated in detail. However, it can be stated that at the moment that no embryonic marker positive germ cells are identified anymore (after 6-12 months), all germ cells present are methylated 10.

13.3 From pluripotency to totipotency of (testicular) germ cell tumors The histological composition of (T)GCTs indicates that these cancers are pluripotent, i.e., can generate derivatives of all three germ cell layers. However, the formation of the germ cell lineage itself was not identified in these tumors so far. In Figure 2 and 3 tumorigenesis and pathogenesis of type II (T)GCTs is illustrated. Using immunohistochemistry with a unique set of recent markers, including OCT3/4, TSPY, and VASA, the presence of germ cell formation was demonstrated in a number of independent nonseminomas 11. Based on these findings, it is concluded that (T)GCTs can also generate the germ cell lineage, making them really toti- or omnipotent. This is in line with the observations that embryonic stem cells are also capable to form the germ cell lineage, both in vitro as well as in vivo 12-14. This supports the use of (T)GCTs and derived cell lines for further investigation of regulation of pluripotency/totipotency, as recently demonstrated 15-16.

13.4 OCT3/4 as ultimate marker for diagnosis and non- invasive detection The capacity of forming the germ cell lineage in individual tumors supports the similarities between pluripotent embryonic stem cells and the cells of origin of (T)GCTs. In this context, OCT3/4 is of particular interest. Originally, OCT3/4 has been identified as regulator of pluripotency in mouse and subsequently human embryonic stem cells 3,17. In addition, it has been found to be expressed in mouse and human PGCs and gonocytes 18. Absence of OCT3/4 in mouse PGCs results in induction of apoptosis 19-21. This indicates that at least an additional, i.e., apoptosis-suppressing role of OCT3/4, exists in PGCs compared to pluripotency regulation in embryonic stem cells. In human germ cells, OCT3/4 is down- regulated upon maturation from gonocyte to pre-spermatogonia (see above), while it remains positive in mice 22-23. This must be kept in mind when comparing different species, especially related to the development of GCTs 24.

190 GENERAL DISCUSSION

Figure 2. Tumorigenesis of the type II (T) GCT. The early PGC stops maturating and is now called a pre- CIS cell which migrates to the genital ridge. After puberty the CIS cell gives rise to an invasive germ cell tumor. SE and/or NSE. The embryonal carcinoma (EC) is the stem cell component of the nonseminomas and can give rise to the different differentiated components: Teratoma (TE), choriocarcinoma (ChC), yolk sac tumor (YST) and germ cell lineage (GCL). SCF is detectable as soon as the PGC becomes arrested and is still present until the CIS transforms to an invasive tumor cell.

Figure 3. Pathogenesis of the type II (T)GCT. The arrested embryonal germ cell (pre-CIS cell) in the embryonal gonad, is supposed to start proliferating at puberty. Via the specific intratubular stage CIS/GB develops in an invasive germ cell tumor. Testicular cancer is the only totipotent cancer that can give rise to extra-somatic tissue, somatic tissues and also, which is for the first time described in chapter 5, to the germ cell lineage.

191 CHAPTER 13

All precursors of (T)GCTs, either CIS of the testis or GB of the dysgenetic gonad, are positive for OCT3/4, without a single exception 3,25-26. The diagnostic utility of this marker can be considered as one of the recent significant contributions to the field of diagnosis of (T)GCTs. Besides the consistent staining of the precursor lesions, OCT3/4 is also positive in all seminomatous tumors, as well as all embryonal carcinomas. This is independent from the primary localization of the cancer, and whether it is metastatic or not. Based on this characteristic, an undiagnosed brain tumor in an adult female was classified as a germinoma (being a seminomatous tumor of the brain), after which an effective treatment approach could be initiated 27. Detection of OCT3/4 can be done using suitable mono- and polyclonal antibodies, as well as after different methods of fixation. However, the use of acid fixatives, especially in case of over-fixation, can negatively effect detection of nuclear markers in general, including OCT3/4 28. The differentiated nonseminomatous elements of (T)GCTs are consistently negative for OCT3/4, both in case of in vivo tumors as well as in vitro cultured and manipulated cell lines. OCT3/4 is therefore the preferred initial diagnostic marker for (T)GCTs, including their precursor lesions. It is found to be more sensitive and specific than c-KIT (CD117) and PLAP 29-33. Of notion is that these latter could not be detected in some case due to poor tissue preservation, while the cancer cells remained positive for OCT3/4. In specific well defined situations OCT3/4 might result in false positive conclusions, which will be discussed later.

The robustness of OCT3/4 as diagnostic marker on tissue sections, including biopsies and cytology specimens, stimulated testing of its applicability as marker in a non-invasive set up. CIS detection in semen of males at risk for development of a TGCT has been an attractive approach previously 34-36, although no real success has been reported. In fact, about 1% of Andrology clinic patients will develop a TGCT, and are therefore a putative population for screening. A pilot study of men at risk for CIS demonstrated that 85% showed OCT3/4 positive cells in semen, of which the presence of CIS was proven based on immunohistochemical methods on biopsy-obtained testicular tissue 37. This positive finding indicates that a non-invasive test for CIS can in principle be developed, of which the characteristics (including positive- and negative predicting power) need to be investigated in ongoing multi-centered studies. A number of questions remain to be answered before this method can be applied in a clinical setting. These include amongst others which patients have to be selected for this procedure, how to deal with negative findings, and about the procedure to follow up in case of a positive finding (e.g. which testis is affected?). In spite of these possible limitations, the approach is of interest to be further developed for clinical implementation. It might be modified in various ways, including the recent possibility to detect protein (i.e., OCT3/4) by means of a

192 GENERAL DISCUSSION quantitative PCR-based approach 16. This might result in an informative non- invasive test for screening of high risk individuals for a TGCT.

13.5 Differential diagnosis of seminomatous tumors and embryonal carcinoma Both the seminomatous tumors as well as embryonal carcinomas are positive for OCT3/4 3,38-40. While the seminomatous cells have a characteristic nuclear staining pattern, embryonal carcinoma cells show both a nuclear and a cytoplasmic localization of the protein 38. In addition, embryonal carcinomas are positive for a number of other markers, like CD30 and keratin, although these can also be sporadically positive in seminoma cells. A distinction between seminoma and embryonal carcinoma is however of patho-biological as well as clinical interest. In fact, seminomas can be treated using irradiation, while this is not used for embryonal carcinomas, which need chemotherapy 41. Therefore availability of a robust diagnostic marker to distinguish seminoma from embryonal carcinoma is of interest. In addition, the presence of intravascular embryonal carcinoma is predictive for development of metastasis in stage I nonseminomas 42-43. Based on a combined mRNA expression profile analysis and protein investigation, a consistent presence of SOX2 in embryonal carcinoma was identified 5. This is in accordance with the knowledge on embryonic stem cells, as well as other studies on TGCTs 44-46. Careful analysis showed that SOX2 is expressed in Sertoli cells within seminiferous tubules with a non-physiological constitution, either being deprived of or with reduced numbers of germ cells, or the presence of CIS. This can be falsely interpreted as intratubular embryonal carcinoma, leading to in fact over-diagnosis 5. Using the same approach, SOX17 was found to be consistently expressed in seminomatous cells, including CIS and GB. However, normal spermatogonia were also positive, therefore SOX17 is not a marker suitable to distinguish normal adult- from malignant germ cells. Interestingly, SOX17 has been reported to be the regulatory transcription factor for the transition between embryonic and adult hematopoietic stem cells 47. As SOX2 can be expressed in more differentiated elements of nonseminomas, SOX17 also shows a wider spectrum of expression. However, the combined positive staining of OCT3/4 with either SOX17 or SOX2, is absolutely informative to identify tumor cells of seminoma and embryonal carcinoma, respectively. It remains to be investigated whether SOX17 forms a similar complex with OCT3/4 as reported for SOX2 48-51, with the same type of transcriptional outcome 5 (for review of transcriptional and signaling networks of pluripotency see 52). Therefore, a limited set of (embryonic) transcription factors, showing a strong and consistent nuclear staining pattern, i.e., OCT3/4, SOX2 and SOX17, is informative for the diagnosis of TGCTs in a routine clinical setting.

193 CHAPTER 13

13.6 SCF as diagnostic marker and early pathogenetic mechanism Although OCT3/4 is an absolute marker for detection of malignant germ cells in the adult testis, it might result in false positive findings in case of delayed maturation of germ cells 53-57. It is known that positive germ cells can be physiologically present during the first months after birth. In Figure 1B the protein expression in aberrant germ cell development is illustrated. However, this time window can be extended in various pathological conditions, including cryptorchidism, and cases of DSD 58-59. These situations are generally characterized by a disturbed testis development, referred to as reduced “testicularization”, associated with a slower or even absent maturation of gonocytes to pre-spermatogonia. Based on various descriptive parameters it has been tried to distinguish the identity of early malignant germ cells from delayed matured germ cells 60. Although this approach is informative to a certain level, it is difficult to be used in a routine diagnostic setting. In this context, the application of immunohistochemistry for SCF (KITLG), is highly informative. No staining has been identified during normal development, while, in contrast, all CIS and GB cases are positive, as well as the high risk samples, showing embryonic germ cells at risk for malignant transformation. This finding is of relevance for a number of reasons. First it suggests a deviation from the normal regulation of SCF in the pathogenesis of (T)GCTs. Suggestive data is available that this is due to activation, or retention, of an autocrine loop, as found in early mouse gonocytes 61. Indeed, a role in migration as well as survival has been reported for this ligand 62-64. The possible impact of the presence of either the membrane bound or soluble form remains to be investigated 65-69. Second, it links this diagnostic marker to genome wide association study results, in which an association between development of a TGCT and various single nucleotide polymorphisms (SNPs) has been found 70-72. All identified SNPs seem to link to genes of which the encoded proteins have a role in generation of the germ cell niche, including SCF. The functional effect of the different SNPs, including those related to SCF, is unknown so far, but highly interesting and critical in elucidation of the pathogenesis of this type of cancer. Of specific value in a clinical setting would be the combined analysis of risk factors, including the SNPs, to develop a predictive model for TGCTs in selected groups of individuals.

194 GENERAL DISCUSSION

13.7 Direct frozen tissue diagnosis of CIS and invasive testicular germ cell tumors Frozen tissue diagnosis can be informative in a clinical context of testicular lesions in particular cases. These relate, amongst others, to the possible limited surgical approach of removing a putative benign lesion (enucleation), like in case of a mono-testis situation. However, detailed investigation of the adjacent parenchyma for the presence of CIS is required to exclude the presence of TGCT. PGCs and gonocytes are positive for PLAP detectable using immunohistochemistry, as well as found for CIS and GB. The enzymatic activity can also be visualized using a direct readout system on frozen tissue, an approach to be performed within a 5 minutes time frame. Accurate application of the method is feasible, due to the fact that endothelial cells of small vessels serve as positive controls. This approach is informative to detect CIS (and GB), as well as seminoma and embryonal carcinoma. It is much faster than standard fixation, paraffin-embedding and subsequent staining using immunohistochemistry, which requires at least 2 days. This so-called dAP test can be used for direct diagnosis during surgical intervention, allowing direct decision making regarding possible limited surgery, preventing possible need for a second operation 73.

13.8 Epigenetics Epigenetic modification is proven to be a crucial regulator of gene expression, both during embryogenesis as well as pathogenesis of various diseases, including cancer 74. One of the modifications is methylation of DNA, of which the overall status can be investigated using immunohistochemistry. Like PGC/gonocytes, CIS and GB have a demethylated pattern, in line with the close relationship between these types of cells. In addition, also the various seminomatous tumor cells are demethylated. In contrast, all differentiated nonseminomatous elements are hypermethylated, while the embryonal carcinoma cells show an intermediate pattern. In spite of a limited number of cases, it is of interest that chemoresistant seminomas show higher methylation. This triggered a study using (T)GCT derived cell lines relating methylation pattern and (cisplatin) chemosensitivity. In spite of the observation that all cell lines show a high level of methylation, different levels of sensitivity to chemotherapy have been reported. Forced demethylation, using 5-azacytidine, in the cisplatin-resistant seminomatous cell line TCam-2 results in induced sensitivity 10. High throughput methylation status investigation identified a number of regulatory sites of genes to be affected by this manipulation. One of the elements is associated with FLAR, also known as c-FLIP. This is a known oncogene, which is inhibiting formation of active caspase 8, which is required for execution of the apoptotic pathway 75. Indeed, overexpression of FLAR is found to suppress apoptosis, also in TGCT-

195 CHAPTER 13 derived cells lines 76, and it is found to be one of the predictive genes in an expression profile analysis of cisplatin-refractory TGCT 77, as well as in other cancers 78-79. These results strongly suggest that apoptosis regulation is one of the determining factors in regulation of treatment sensitivity of TGCTs. Indeed, such a link has been reported before 76,80. Further investigations are required to understand the impact of this finding in a clinical setting. For example, a combined approach, in which demethylating agents have a place, could be investigated in the treatment of cisplatin-refractory (T)GCTs.

13.9 Concluding remarks Based on the studies performed, a number of conclusions can be drawn. Optimal diagnosis of (T)GCT, including their precursor lesions CIS and GB, is feasible using a limited set of markers in a routine clinical setting. This approach, including antibodies against OCT3/4, SOX17, SOX2 and SCF, is applicable independent of the age of the patient, as well as localization of the cancer. Moreover, an absolute test for the presence of CIS on frozen tissue is developed, suitable to be included into direct clinical decision making for limited surgical procedures. The overall and targeted methylation status is found to be involved in treatment sensitivity and resistance. FLAR is an interesting target for further investigation. Combining the observation of SCF as early diagnostic tool and the predictive power of SNPs related to this gene, opens a complete new field on investigation on the pathogenesis of (T)GCTs. In this context the interaction between Genetics and Environment, referred to as GENVIRONMENT, is an exciting development to be further explored in the context of the pathogenesis of (T)GCTs.

References 1. Honecker, F., et al. New insights into the pathology and molecular biology of human germ cell tumors. World J Urol 22, 15-24 (2004). 2. Stoop, H., et al. Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological implications. Lab Invest 81, 919-928 (2001). 3. Looijenga, L.H., et al. POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Cancer Res 63, 2244-2250 (2003). 4. Stoop, H., et al. Stem cell factor as a novel diagnostic marker for early malignant germ cells. J Pathol 216, 43-54 (2008). 5. de Jong, J., et al. Differential expression of SOX17 and SOX2 in germ cells and stem cells has biological and clinical implications. J Pathol 215, 21-30 (2008). 6. Gaskell, T.L., Esnal, A., Robinson, L.L., Anderson, R.A. & Saunders, P.T. Immunohistochemical profiling of germ cells within the human fetal testis: identification of three subpopulations. Biol Reprod 71, 2012-2021 (2004). 7. Kerr, C.L., Hill, C.M., Blumenthal, P.D. & Gearhart, J.D. Expression of pluripotent stem cell markers in the human fetal testis. Stem Cells 26, 412-421 (2008). 8. Sonne, S.B., et al. Analysis of gene expression profiles of microdissected cell populations indicates that testicular carcinoma in situ is an arrested gonocyte. Cancer Res 69, 5241- 5250 (2009).

196 GENERAL DISCUSSION

9. De Miguel, M.P., Fuentes-Julian, S. & Alcaina, Y. Pluripotent stem cells: origin, maintenance and induction. Stem Cell Rev 6, 633-649 (2010). 10. Wermann, H., et al. Global DNA methylation in fetal human germ cells and germ cell tumours: association with differentiation and cisplatin resistance. J Pathol 221, 433-442 (2010). 11. Honecker, F., et al. Germ cell lineage differentiation in non-seminomatous germ cell tumours. J Pathol 208, 395-400 (2006). 12. Toyooka, Y., Tsunekawa, N., Akasu, R. & Noce, T. Embryonic stem cells can form germ cells in vitro. Proc Natl Acad Sci U S A 100, 11457-11462 (2003). 13. Geijsen, N., et al. Derivation of embryonic germ cells and male gametes from embryonic stem cells. Nature 427, 148-154 (2004). 14. Clark, A.T., et al. Spontaneous differentiation of germ cells from human embryonic stem cells in vitro. Hum Mol Genet 13, 727-739 (2004). 15. Geijsen, N. & Jones, D.L. Seminal discoveries in regenerative medicine: contributions of the male germ line to understanding pluripotency. Hum Mol Genet 17, R16-22 (2008). 16. Gillis, A.J., et al. Expression and interdependencies of pluripotency factors LIN28, OCT3/4, NANOG and SOX2 in human testicular germ cells and tumours of the testis. Int J Androl (2011). 17. Surani, M.A., Hayashi, K. & Hajkova, P. Genetic and epigenetic regulators of pluripotency. Cell 128, 747-762 (2007). 18. Loh, Y.H., et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat Genet 38, 431-440 (2006). 19. Nichols, J., et al. Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4. Cell 95, 379-391 (1998). 20. Ginis, I., et al. Differences between human and mouse embryonic stem cells. Dev Biol 269, 360-380 (2004). 21. Kehler, J., et al. Oct4 is required for primordial germ cell survival. EMBO Rep 5, 1078-1083 (2004). 22. Ovitt, C.E. & Scholer, H.R. The molecular biology of Oct-4 in the early mouse embryo. Mol Hum Reprod 4, 1021-1031 (1998). 23. Gidekel, S., Pizov, G., Bergman, Y. & Pikarsky, E. Oct-3/4 is a dose-dependent oncogenic fate determinant. Cancer Cell 4, 361-370 (2003). 24. Looijenga, L.H., Gillis, A.J., Stoop, H.J., Hersmus, R. & Oosterhuis, J.W. Chromosomes and expression in human testicular germ-cell tumors: insight into their cell of origin and pathogenesis. Ann N Y Acad Sci 1120, 187-214 (2007). 25. de Jong, J., et al. Diagnostic value of OCT3/4 for pre-invasive and invasive testicular germ cell tumours. J Pathol 206, 242-249 (2005). 26. Kersemaekers, A.M., et al. Identification of germ cells at risk for neoplastic transformation in gonadoblastoma. Hum Pathol 36, 512-521 (2005). 27. de Jong, J., et al. A 40-year-old woman with a progressive periventricular white matter lesion. Brain Pathol 18, 103-104, 142 (2008). 28. van Casteren, N.J., et al. Evaluation of testicular biopsies for carcinoma in situ: immunohistochemistry is mandatory. Int J Androl 32, 666-674 (2009). 29. Izquierdo, M.A., et al. Differential expression of the c-kit proto-oncogene in germ cell tumours. J Pathol 177, 253-258 (1995). 30. Rajpert-De Meyts, E., Kvist, M. & Skakkebæk, N.E. Heterogeneity of expression of immunohistochemical tumour markers in testicular carcinoma in situ: pathogenetic relevance. Virchows Arch 428, 133-139 (1996). 31. Jones, T.D., Ulbright, T.M., Eble, J.N. & Cheng, L. OCT4: A sensitive and specific biomarker for intratubular germ cell neoplasia of the testis. Clin Cancer Res 10, 8544-8547 (2004).

197 CHAPTER 13

32. Hattab, E.M., Tu, P.H., Wilson, J.D. & Cheng, L. OCT4 Immunohistochemistry Is Superior to Placental Alkaline Phosphatase (PLAP) in the Diagnosis of Central Nervous System Germinoma. Am J Surg Pathol 29, 368-371 (2005). 33. Emerson, R.E. & Ulbright, T.M. Intratubular germ cell neoplasia of the testis and its associated cancers: the use of novel biomarkers. Pathology 42, 344-355 (2010). 34. Giwercman, A., Hopman, A.H., Ramaekers, F.C. & Skakkebæk, N.E. Carcinoma in situ of the testis. Detection of malignant germ cells in seminal fluid by means of in situ hybridization. Am J Pathol 136, 497-502 (1990). 35. Meng, F.J., Zhou, Y., Skakkebæk, N.E., Marks, A. & Giwercman, A. Detection and enrichment of carcinoma-in-situ cells in semen by an immunomagnetic method using monoclonal antibody M2A. Int J Androl 19, 365-370 (1996). 36. Hoei-Hansen, C.E., et al. Towards a non-invasive method for early detection of testicular neoplasia in semen samples by identification of fetal germ cell-specific markers. Hum Reprod 22, 167-173 (2007). 37. van Casteren, N.J., et al. Noninvasive detection of testicular carcinoma in situ in semen using OCT3/4. Eur Urol 54, 153-158 (2008). 38. de Jong, J., et al. Diagnostic value of OCT3/4 for pre-invasive and invasive testicular germ cell tumours. J Pathol (2005). 39. Skotheim, R.I., et al. Differentiation of human embryonal carcinomas in vitro and in vivo reveals expression profiles relevant to normal development. Cancer Res 65, 5588-5598 (2005). 40. Iczkowski, K.A., et al. Trials of new germ cell immunohistochemical stains in 93 extragonadal and metastatic germ cell tumors. Hum Pathol 39, 275-281 (2008). 41. Albers, P., et al. EAU Guidelines on Testicular Cancer: 2011 Update. Eur Urol 60, 304-319 (2011). 42. Foster, R.S. & Nichols, C.R. Testicular cancer: what's new in staging, prognosis, and therapy. Oncology (Williston Park) 13, 1689-1694; discussion 1697-1700, 1703 (1999). 43. Stephenson, A.J. & Klein, E.A. Surgical management of low-stage nonseminomatous germ cell testicular cancer. BJU Int 104, 1362-1368 (2009). 44. Boyer, L.A., et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 122, 947-956 (2005). 45. Biermann, K., et al. Gene expression profiling identifies new biological markers of neoplastic germ cells. Anticancer Res 27, 3091-3100 (2007). 46. Korkola, J.E., et al. In vivo differentiation and genomic evolution in adult male germ cell tumors. Genes Chromosomes Cancer 47, 43-55 (2008). 47. Kim, I., Saunders, T.L. & Morrison, S.J. Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem cells. Cell 130, 470-483 (2007). 48. Remenyi, A., et al. Crystal structure of a POU/HMG/DNA ternary complex suggests differential assembly of Oct4 and Sox2 on two enhancers. Genes Dev 17, 2048-2059 (2003). 49. Catena, R., et al. Conserved POU binding DNA sites in the Sox2 upstream enhancer regulate gene expression in embryonic and neural stem cells. J Biol Chem 279, 41846-41857 (2004). 50. Chew, J.L., et al. Reciprocal transcriptional regulation of Pou5f1 and Sox2 via the Oct4/Sox2 complex in embryonic stem cells. Mol Cell Biol 25, 6031-6046 (2005). 51. Okumura-Nakanishi, S., Saito, M., Niwa, H. & Ishikawa, F. Oct-3/4 and Sox2 regulate Oct- 3/4 gene in embryonic stem cells. J Biol Chem 280, 5307-5317 (2005). 52. Ng, H.H. & Surani, M.A. The transcriptional and signalling networks of pluripotency. Nat Cell Biol 13, 490-496 (2011). 53. Rajpert-De Meyts, E., Jorgensen, N., Brondum-Nielsen, K., Muller, J. & Skakkebæk, N.E. Developmental arrest of germ cells in the pathogenesis of germ cell neoplasia. Apmis 106, 198-204; discussion 204-196 (1998). 54. Rajpert-De Meyts, E., et al. The emerging phenotype of the testicular carcinoma in situ germ cell. Apmis 111, 267-278; discussion 278-269 (2003). 198 GENERAL DISCUSSION

55. Honecker, F., et al. Pathobiological implications of the expression of markers of testicular carcinoma in situ by fetal germ cells. J Pathol 203, 849-857 (2004). 56. Cools, M., Drop, S.L., Wolffenbuttel, K.P., Oosterhuis, J.W. & Looijenga, L.H. Germ cell tumors in the intersex gonad: old paths, new directions, moving frontiers. Endocr Rev 27, 468-484 (2006). 57. Cools, M., et al. Maturation delay of germ cells in fetuses with trisomy 21 results in increased risk for the development of testicular germ cell tumors. Hum Pathol 37, 101-111 (2006). 58. Looijenga, L.H., et al. Tumor risk in disorders of sex development (DSD). Best Pract Res Clin Endocrinol Metab 21, 480-495 (2007). 59. Pleskacova, J., et al. Tumor risk in disorders of sex development. Sex Dev 4, 259-269 (2010). 60. Cools, M., et al. Morphological and immunohistochemical differences between gonadal maturation delay and early germ cell neoplasia in patients with undervirilization syndromes. J Clin Endocrinol Metab 90, 5295-5303 (2005). 61. Yabuta, Y., Kurimoto, K., Ohinata, Y., Seki, Y. & Saitou, M. Gene expression dynamics during germline specification in mice identified by quantitative single-cell gene expression profiling. Biol Reprod 75, 705-716 (2006). 62. Besmer, P., et al. The kit-ligand (steel factor) and its receptor c-kit/W: pleiotropic roles in gametogenesis and melanogenesis. Dev Suppl, 125-137 (1993). 63. Strohmeyer, T., et al. Expression of the c-kit proto-oncogene and its ligand stem cell factor (SCF) in normal and malignant human testicular tissue. J Urol 153, 511-515 (1995). 64. Mauduit, C., Hamamah, S. & Benahmed, M. Stem cell factor/c-kit system in spermatogenesis. Hum Reprod Update 5, 535-545 (1999). 65. Godin, I., et al. Effects of the steel gene product on mouse primordial germ cells in culture. Nature 352, 807-809 (1991). 66. Dolci, S., et al. Requirement for mast cell growth factor for primordial germ cell survival in culture. Nature 352, 809-811 (1991). 67. Matsui, Y., et al. Effect of Steel factor and leukaemia inhibitory factor on murine primordial germ cells in culture. Nature 353, 750-752 (1991). 68. Marziali, G., Lazzaro, D. & Sorrentino, V. Binding of germ cells to mutant Sld Sertoli cells is defective and is rescued by expression of the transmembrane form of the c-kit ligand. Dev Biol 157, 182-190 (1993). 69. Broudy, V.C. Stem cell factor and hematopoiesis. Blood 90, 1345-1364 (1997). 70. Rapley, E. Susceptibility alleles for testicular germ cell tumour: a review. Int J Androl 30, 242-250; discussion 250 (2007). 71. Kanetsky, P.A., et al. Common variation in KITLG and at 5q31.3 predisposes to testicular germ cell cancer. Nat Genet 41, 811-815 (2009). 72. Turnbull, C., et al. Variants near DMRT1, TERT and ATF7IP are associated with testicular germ cell cancer. Nat Genet 42, 604-607 (2010). 73. Stoop, H., et al. Diagnosis of testicular carcinoma in situ'(intratubular and microinvasive)' seminoma and embryonal carcinoma using direct enzymatic alkaline phosphatase reactivity on frozen histological sections. Histopathology 58, 440-446 (2011). 74. Lotem, J. & Sachs, L. Epigenetics and the plasticity of differentiation in normal and cancer stem cells. Oncogene 25, 7663-7672 (2006). 75. Boatright, K.M. & Salvesen, G.S. Mechanisms of caspase activation. Curr Opin Cell Biol 15, 725-731 (2003). 76. Spierings, D.C., de Vries, E.G., Vellenga, E. & de Jong, S. The attractive Achilles heel of germ cell tumours: an inherent sensitivity to apoptosis-inducing stimuli. J Pathol 200, 137-148 (2003). 77. Korkola, J.E., et al. Identification and validation of a gene expression signature that predicts outcome in adult men with germ cell tumors. J Clin Oncol 27, 5240-5247 (2009).

199 CHAPTER 13

78. Lima, R.T., et al. MicroRNA regulation of core apoptosis pathways in cancer. Eur J Cancer 47, 163-174 (2011). 79. Safa, A.R., Day, T.W. & Wu, C.H. Cellular FLICE-like inhibitory protein (C-FLIP): a novel target for cancer therapy. Curr Cancer Drug Targets 8, 37-46 (2008). 80. Mueller, T., et al. Failure of activation of caspase-9 induces a higher threshold for apoptosis and cisplatin resistance in testicular cancer. Cancer Res 63, 513-521 (2003).

200

A:IM>KØÛ

&H@@4EL &4@8AI4GG

CHAPTER 14

English

202 SUMMARY/SAMENVATTING

Introduction The aim of the work described in this thesis is to improve the understanding of the pathobiology of testicular cancer (type II Germ Cell Tumors) to create possibilities for optimalization of diagnosis for this type of malignancy in routine pathology laboratories. The different studies presented here show valuable additional information on the microscopic diagnostics in daily practice. This enables proper and complete diagnosis of this relative rare variant of cancer ensuring the best possible treatment for the patient. A number of novel sensitive and specific immunohistochemical markers are presented, of value for (early) diagnostics. In addition an enzymhistochemical method is presented informative to detect the presence of (pre-)malignant cells during a surgical intervention. Based on these data, a decision can be made regarding a testis sparing procedure. The studies included in this thesis are part of a diagnostically decision tree in which immuno- and enzymhistochemistry plays an important role.

General Testicular cancer is a relative rare variant of cancer with raising incidence in the Western population, in contrast to the rest of the world. Within the Netherlands, about 750 new patients are diagnosed each year. In particular this affects young males between the age of 15 – 45 years The current treatment possibilities is successful in approximately 95% of the patients. It consists of removing of the affected testicle (radical orchiectomy) and possibly additional treatment using irradiation and/or chemotherapy. In spite of the excellent prognosis, patients often will suffer from long-term side effects of the treatment like fatigue, disorders of the metabolism, heart diseases, neurotoxicity and infertility. A small group of patients (5%) will shows an insensitivity of the cancer treatment given, and will possibly die of the disease. An early and accurate diagnosis can prevent under- or overtreatment, ensure an optimal balance between the pro and conts of the treatment. This will have significant impact on their quality of life, for both short- and long term.

Specific The cell of origin from which testicular cancer arises is already present during the first months of pregnancy (embryogenesis), as demonstrated in Chapter 3. It is a primitive germ cell formed after the first month after fertilization, which provide spermatozoa after puberty by spermatogenesis. In case this primitive germ cell does not maturate completely and keeps the properties of the embryonal cell. This cell can become carcinoma in situ (CIS), being the progenitor of testicular cancer. These CIS cell is located in the testis on the same place as the spermatogonium, below the Sertoli cells on the basal lamina. These cells start to

203 CHAPTER 14 multiply in an uncontrolled manner after puberty and finally they can develop cancer. In chapter 4 the totipotent character of this cancer is described for the first time in literature, shown by the presence of new formed germ cells. In other words testicular cancer is the only real totipotent cancer that can give rise to different germ lines the same as in embryogenesis. The application of a unique panel of immunological markers permitted this finding. In chapter 5 the value of the presence of OCT3/4 in CIS, seminoma and embryonal carcinoma is for the first time described. This study was initiated by the presence of this marker in embryonal cells and germ cells of the mouse. The diagnostic significance of this extremely sensitive and specific marker is huge as described in chapter 6. Diagnostics of testicular germ cell tumors requires in most cases surgical removed tissue. In chapter 7 a new method is described that offers the possibilities to detect OCT3/4 positive cells in semen. This method introduces the possibility to screen men with a higher risk for developing testicular germ cell tumor for CIS cells without surgical intervention. The robustness of this detection method for OCT3/4, even in sub optimal material, is illustrated in chapter 8. The malignant germ cells in the here described patient where only detectable by the presence of OCT3/4.

The two groups of testicular cancer: seminomatous and nonseminomatous tumors, differs in sensitivity for irradiation and chemotherapy. Therefore differentiation of these two groups is of great importance for an optimal treatment. The nonseminomas can be subdivided in different groups of which embryonal carcinoma is the stem cell population. In chapter 9 detection of two SOX proteins (SOX2 and SOX17) is shown to be informative. Seminoma cells show expression of OCT3/4 and SOX17, embryonal carcinoma cells of OCT3/4 and SOX2. The other subgroups of the nonseminomas (yolk sac tumor, choriocarcinoma and teratoma) show a more heterogeneous expression.

A chance of overdiagnosis of CIS in the testicle based on the presence of OCT3/4 is risk in young children, especially in the first year after birth, a result of a possible delay in maturation of the normal developing germ cells. This is Important because maturation delay is common also in cases of an increased risk for the development of testicular cancer, for instance an undescended testicle. In chapter 10 it is demonstrated how to distinguish a germ cell with maturation delay from a (pre-) malignant germ cell (CIS). Because SCF is only present in the direct neighbourhood of (pre-)malignant germ cells. This finding has an additional value in case of possible maturation delay when OCT3/4 positive cells are present. There is also a significant value for this finding concerning that specific DNA variants related to this gene, associated with an increased risk for development of testicular cancer in the entire population.

204 SUMMARY/SAMENVATTING

Organ sparing surgery may be preferred for different reasons, even so in cases of testicular cancer. Prevention of a second surgical intervention requires proper histopathological examination of frozen sections. In testis cancer this is mainly based on the detection of the presence of CIS. Unfortunately morphologic analysis of frozen material is not that easy, therefore formalin fixed tissue is needed which excludes often a single operation procedure. In chapter 11 a direct enzymhistochemical method is described which makes it possible to detect the presence of these CIS cells in a few minutes on frozen tissue. By using this direct alkaline phosphatase (dAP) technique it is possible to have certainty about the presence of CIS cells in the available tissue. Based on these results a decision is possible about performing a testis sparing operation.

Chapter 12 describes the low DNA methylation status of both CIS cells and embryonal germ cells in their natural environment. This is performed by the use of immunohistochemistry. Interesting is the possible connection between cancer cells which are sensitive to cisplatin and the status of methylation. DNA of the resistant tumors seems to be hypermethylated compared to sensitive tumors. Analysis of cultured cell lines indicates c-FLIP gene as a potential candidate. This gene regulates the sensitivity of cells for a programmed cell death (apoptosis).

Conclusions The studies presented give possibilities to perform optimal diagnostics of testicular cancer, eventually in certain circumstances, even in a non-invasive set up. These findings demonstrate that understanding of the normal development of germ cells, especially during embryogenesis, is of importance to carry out these kind of studies. The results give a solid base for implementation in a diagnostic set up and for further translational research.

205 CHAPTER 14

Nederlands

206 SUMMARY/SAMENVATTING

Introductie Het doel van het werk beschreven in dit proefschrift is het begrijpen van de pathobiologie van zaadbalkanker (teelbalkanker, kiemcelkanker), en daarmee mogelijkheden te creëren voor het optimaliseren van de diagnostiek van deze vorm van maligniteit, toepasbaar in elk pathologisch laboratorium. De hier beschreven studies geven een waardevolle aanvulling op de microscopische diagnostiek in de dagelijkse praktijk. Hierdoor is een betrouwbare en volledige diagnose te stellen van deze relatief zeldzame vorm van kanker die voor de behandeling van de patiënt van belang is.

Er worden een aantal nieuwe gevoelige en specifieke immunologische markers geïntroduceerd die voor de (vroeg)diagnostiek een belangrijke toegevoegde waarde bezitten. Tevens wordt een enzymhistochemische methode beschreven die de mogelijkheid biedt om op een eenvoudige toe te passen wijze gedurende een chirurgische ingreep aan de teelbal de aanwezigheid van (pre)maligne cellen te detecteren. Op grond van de bevindingen kan besloten worden tot eventueel een testis-sparende ingreep uit te voeren. De in dit proefschrift beschreven studies maken deel uit van een diagnostische beslisboom waarin immuno- en enzymhistochemie een belangrijke plaats innemen.

Algemeen Zaadbalkanker is een relatief zeldzame vorm van kanker die steeds meer voorkomt in de Westerse populatie, in tegenstelling tot in de rest van de wereld. In Nederland wordt bij ongeveer 750 nieuwe patiënten per jaar deze diagnose gesteld. Vooral jonge mannen tussen de 15 en 45 jaar worden door deze vorm van kanker getroffen.

De huidige beschikbare behandelingsmethoden van zaadbalkanker leidt bij ongeveer 95% van de patiënten tot complete genezing. Dit bestaat uit verwijdering van de aangedane zaadbal (orchidectomie) en eventueel een nabehandeling met bestraling en/of chemotherapie. Ondanks de uitstekende prognose krijgen deze patiënten vaak te maken met de gevolgen van de behandeling die kunnen bestaan uit: vermoeidheid, stofwisselingsstoornissen, hart- en vaat ziekten, aandoeningen aan de zenuwen en onvruchtbaarheid. In ongeveer 5% van de patiënten blijkt tijdens de behandeling de kanker niet gevoelig te zijn, waardoor de patiënt de kans loopt aan de ziekte te overlijden. Een vroege en juiste diagnose kan onder- en overbehandeling voorkomen, waardoor de optimale balans tussen voor- en nadelen van de behandeling verkregen kan worden. Dit heeft significante effecten op de kwaliteit van leven, zowel op korte als op lange termijn.

207 CHAPTER 14

Specifiek De cel van oorsprong waaruit de zaadbalkanker ontstaat is al tijdens de eerste maanden van de zwangerschap (embryogenese) aanwezig, zoals beschreven in Hoofdstuk 3. Het is een primitieve kiemcel die na de eerste maand na de bevruchting gevormd wordt, welke na de pubertijd door middel van spermatogenese bij de man de uiteindelijke zaadcellen gaat vormen. In het geval dat deze primitieve kiemcel niet volledig uitrijpt en embryonale eigenschappen behoudt, kan deze cel carcinoma in situ (CIS) vormen, dat de voorloper is van zaadbalkanker. De CIS cel bevindt zich in de zaadbal op de plaats van normaal gesproken een zogenaamd spermatogonium zich bevindt (onder de Sertoli cellen, op de basale lamina). Het zijn deze cellen die zich na het starten van de pubertijd ongecontroleerd gaan vermeerderen, en uiteindelijk de kanker kunnen vormen. In hoofdstuk 4 wordt voor de eerste maal in de literatuur de werkelijke totipotentie van zaadbalkanker beschreven, aangetoond door de aanwezigheid van de nieuw gevormde kiemcellen in de kanker. Met andere woorden, zaadbalkanker is de enige werkelijke totipotente kanker, in staat tot het vormen van alle verschillende differentiatielijnen die gevonden kunnen worden tijdens de embryogenese. Deze bevinding is mogelijk gemaakt door de toepassing van een unieke set van immunologische markers. In Hoofdstuk 5 wordt voor de eerste keer de waarde van de aanwezigheid van OCT3/4 voor het aantonen van CIS, seminoom en embryonaal carcinoom beschreven. De reden voor het initiëren van deze studie was de beschrijving van deze marker voor embryonale stam cellen en kiemcellen in de muis. De diagnostische waarde van deze zeer gevoelige en specifieke marker is groot, zoals in hoofdstuk 6 is weergegeven. Voor de diagnostiek van zaadbalkanker wordt in de meeste gevallen gebruik gemaakt van chirurgisch verwijderd weefsel. In hoofdstuk 7 wordt een nieuwe methode beschreven die het mogelijkheid maakt om OCT3/4 positieve cellen in zaadvloeistof aan te tonen. Deze methode biedt de mogelijkheid om bij mannen met een verhoogd risico op het krijgen van zaadbalkanker, zonder een eerste invasieve ingreep, onderzoek te doen naar de aanwezigheid van CIS cellen. De robuustheid van de methode voor het aantonen van OCT3/4 wordt aangetoond, zelfs in sub-optimaal materiaal. Dit wordt geïllustreerd in hoofdstuk 8, waar een patiënt wordt beschreven waarbij de kwaadaardige kiemcellen enkel te detecteren waren door het aantonen van OCT3/4.

Seminomen en nonseminomen zijn twee varianten van zaadbalkanker, die verschillende gevoeligheden vertonen voor bestraling en chemotherapie. Voor het toepassen van de optimale behandeling is een differentiatie dus van belang. Binnen de nonseminomen worden verschillende subvarianten beschreven, waarvan embryonaalcarcinoom de stam cel populatie is. In hoofdstuk 9 wordt beschreven dat het aantonen van twee verschillende SOX eiwitten en wel SOX2

208 SUMMARY/SAMENVATTING en SOX17, hier informatief voor is. De seminoom cellen zijn positief voor OCT3/4 en SOX17 en de embryonaal carcinoomcellen voor OCT3/4 en SOX2. Een meer heterogeen beeld wordt gevonden in de andere nonseminoom componenten (dooierzaktumor, choriocarcinoom en teratoom).

De mogelijke overdiagnostiek van CIS van de zaadbal op grond van het aantonen van OCT3/4 is een risico bij jonge kinderen, met name in het eerste levensjaar. Dit is het gevolg van mogelijke rijpingsremming (vertraagde ontwikkeling) van de normale kiemcellen. Dit is van belang, omdat rijpingsremming voorkomt in die gevallen waarbij ook een verhoogd risico op zaadbalkanker gevonden wordt, zoals bijvoorbeeld een niet ingedaalde zaadbal. In hoofdstuk 10 wordt aangetoond dat er onderscheid gemaakt kan worden tussen een kiemcel met rijpingsremming aan de ene kant en een (pre-)maligne kiemcel (CIS) aan de andere kant. Dit op grond van het aantonen van de aanwezigheid van SCF, dat alleen voorkomt in de directe omgeving van (pre-)maligne kiemcellen. Deze bevinding heeft additionele waarde in het geval dat OCT3/4 positieve kiemcellen aangetoond worden, waarbij de vraag is of er een vertraagde uitrijping aanwezig is. Deze bevinding heeft tevens significante waarde in het kader van het feit dat specifieke DNA varianten gerelateerd aan dit gen beschreven zijn, die een verhoogd risico geven op zaadbalkanker in de totale populatie.

Orgaansparende operaties kunnen op grond van verschillende redenen van belang zijn, zo ook in het geval van zaadbalkanker. Het voorkomen van het nodig zijn van een tweede ingreep vereist informatieve vriescoupe diagnostiek. In het geval van zaadbalkanker betreft dit voornamelijk het aantonen van de aanwezigheid van CIS. Echter blijkt dit door morfologische analyse van vriesmateriaal niet eenvoudig te zijn, waardoor formaline-gefixeerd weefsel nodig is. Hierdoor blijkt de ingreep in een eenmalige chirurgische uitvoering vaak niet mogelijk te zijn. In hoofdstuk 11 wordt een directe enzymhistochemische bepaling beschreven, dat het mogelijk maakt om binnen enkele minuten te bepalen of CIS cellen aanwezig zijn. Met gebruikmaking van deze directe alkalische fosfatase (dAP) test is het mogelijk om met zekerheid vast te stellen of CIS cellen in het beschikbare weefsel aanwezig is. Op grond hiervan is een besluit omtrent het uitvoeren van een zaadbalsparende operatie mogelijk.

In hoofdstuk 12 wordt aangetoond dat de CIS cellen, net als de embryonale kiemcellen een laag niveau van DNA methylatie bevatten. Dit is bepaald door middel van immunohistochemie, waardoor de informatie verkregen kan worden van cellen in hun natuurlijke omgeving. Interessant is de bevinding dat er een mogelijk verband bestaat tussen cisplatin gevoeligheid van de kankercellen en de methylatie status. Met name de ongevoelige tumoren blijken meer DNA methylatie te bevatten dan de gevoelige tumoren. Door cellijn analyse is hierbij

209 CHAPTER 14 het c-FLIP gen als mogelijke kandidaat naar voren gekomen. Dit gen reguleert de gevoeligheid van cellen voor een geprogrammeerde celdood (apoptose).

Conclusies De beschreven bevindingen geven de mogelijkheid optimale diagnostiek van zaadbalkanker uit te voeren, eventueel in een niet-invasieve opzet in bepaalde situaties. De bevindingen maken duidelijk dat het begrijpen van de normale ontwikkeling van kiemcellen, met name tijdens de embryogenese, van belang is voor het uit kunnen voeren van dergelijke studies. De resultaten geven een goede basis voor implementatie in een diagnostische setting als ook voor verder translationeel onderzoek.

210 Dankwoord Mijn ouders, mijn vrouw en kinderen verdienen hier om als eerste genoemd te worden. Sjaak † en Lisa aan jullie heb ik te danken dat ik besta en Els, nooit heb je aan me getwijfeld en altijd gezegd dat ik datgene moet doen waar ik in geloof. Bedankt voor dit prima advies. Mijn dochters Jackelien en Ingrid die direct bereid waren om tijdens de openbare verdediging mijn paranimfen te willen zijn en me geholpen hebben bij het ontwerpen van de lay-out van dit proefschrift. Meiden ik ben trots op jullie en niet alleen hierom.

Leendert en Wolter, bedankt voor het vertrouwen en de waardering die jullie me hebben gegeven vanaf de eerste dag dat ik voor jullie ben komen werken. Dit proefschrift is een direct gevolg hiervan. Ik heb heel veel van jullie geleerd en ben ervan overtuigd dat nog geruime tijd te blijven doen. Mijn collega’s Ad en Remko (met een K) en ex-collega Ruud bedankt voor de samenwerking. Mede door jullie heb ik veel plezier in mijn werk. Vooral de duidelijke, soms voorspelbare, aanwezigheid van Ad geeft altijd weer een reden tot hilariteit evenals het imitatietalent van Remko. Leuk om dat elke dag weer mee te maken. Katja, Willem, Jos en Linda, bedankt voor het aanleveren van onderzoeksmateriaal en de daar aan gekoppelde specifieke vragen en wensen. Dit zorgt ervoor dat ik aardig bezig kan blijven.

In de jaren dat ik werkzaam ben in de groep van Leendert heb ik met veel aio’s, oio’s en hoe ze verder ook mogen heten, samengewerkt en altijd met plezier. Ik zal er een aantal hier noemen. Ben ik je vergeten, dan is dat zeker niet bewust gebeurd. De eerste was Marijke die tijdens haar promotie al tegen me zei: “jij bent de volgende”. Het heeft wat langer geduurd, maar uiteindelijk is het er dan toch van gekomen. Mijn Duitse vrienden Frank, Friedemann en Hendrik, Martine uit mijn favoriete stad Antwerpen, Gaetano en Claudia uit Italië, Michel, Jeroen, Niels, Jana, Bestari, Martine ook oorspronkelijk uit België, Berdine, Charlotte, Ton, Marsha, Ronak en de altijd kritische Lambert, erg fijn om jullie te hebben leren kennen en ik vind (vond) het plezierig om met jullie te mogen samen werken. Een aparte plaats hier is voor Wil die altijd een grote belangstelling heeft voor mijn moeder, mijn vrouw, mijn kinderen en kleinkinderen en die in de jaren dat ze bij ons werkt al heel wat ervaring heeft gekregen in het werken met een computer. Wil, nog bedankt voor je hulp. Martin wil ik speciaal bedanken voor de enorme hulp bij het samenstellen en in elkaar zetten van dit proefschrift. Ik denk niet dat ik het zonder hem zo snel voor elkaar had gehad.

Zonder de hulp van het histo- en immunolab. was het een stuk lastiger geweest om mijn werk goed te kunnen uitvoeren. Allemaal reuze bedankt. En ook Frank voor de hulp bij allerlei problemen omtrent de fotografie, zowel privé als werk gerelateerd.

211 Ten slotte wil ik de MMV hartelijk danken voor hun ondersteuning bij dit proefschrift. Ik draag deze club al heel wat jaren een warm hart toe. Een speciaal woord van dank voor mijn vrienden drs. de Graaf (Simon) † en Cor, met wie ik al vele jaren deel uit maak van de redactiecommissie van het tijdschrift Uitzicht van de MMV. Elke vergadering vroeg Simon weer wanneer ik ging promoveren. Jammer dat je er nu niet bij kunt zijn, Simon. “Ik ben dan op vakantie”, zei je toen ik je vertelde dat het eindelijk ging gebeuren. De introductie heb ik je gelukkig nog kunnen laten lezen.

212

Curriculum Vitae The author of this thesis was born in Bergen op Zoom on December 1st, 1955. In 1975 he finished his HBO-A pathology, histology, cytology and hematology at the School voor Laboratoriumpersoneel in Breda and started his career as a technician at the laboratory for histo-pathology ( department of pathology ) at the University Hospital Rotterdam. After 3 years he became a chief technician. In 1982 he started as a chief technician of the department of immunohistochemistry at the Pathological Anatomical Laboratory in Dordrecht. In June 1994 he graduated from the Polytechnische Faculteit Hogeschool West Brabant in Etten Leur: HLO-Biochemistry and started to study biology at the University Utrecht in the same year. During this period he finished his internship molecular genetics with the dissertation “The instability of chromosomes by an aberrant mitotic checkpoint in human testicular germ cell tumors” at the Laboratory for Experimental Pathology and Oncology ( L.E.P.O. ) of Prof.dr. Leendert Looijenga and Prof.dr. Wolter Oosterhuis at the Daniel den Hoed Cancer Center Rotterdam. In 1998 he graduated at the University Utrecht specialized in “Molecular Genetics” and “History and Philosophy of Biology”. Since January 1999 he has been working as a senior research technician with the group of Prof.dr. Leendert Looijenga at the Josephine Nefkens Institute, Erasmus Medical Center Rotterdam. He is married to Els and they have two daughters and two grandsons.

213 List of publications o Specific detection of OCT3/4 isoform A/B/B1 expression in solid (germ cell) tumours and cell lines:confirmation of OCT3/4 specificity for germ cell tumours. Rijlaarsdam MA, van Herk HA, Gillis AJ, Stoop H, Jenster G, Martens J, van Leenders GJ, Dinjens W,Hoogland AM, Timmermans M, Looijenga LH. Br J Cancer. 2011 Sep 6;105(6):854-63 o A pathologist's view on the testis biopsy. Oosterhuis JW, Stoop H, Dohle G, Boellaard W, van Casteren N, Wolffenbuttel K, Looijenga LH. Int J Androl. 2011 Aug;34(4 o Expression and interdependencies of pluripotency factors LIN28, OCT3/4, NANOG and SOX2 in human testicular germ cells and tumours of the testis. Gillis AJ, Stoop H, Biermann K, van Gurp RJ, Swartzman E, Cribbes S, Ferlinz A, Shannon M, Wolter Oosterhuis J, Looijenga LH. Int J Androl. 2011 Jun 2. o Dissecting the molecular pathways of (testicular) germ cell tumour pathogenesis; from initiation to treatment-resistance. Looijenga LH, Gillis AJ, Stoop H, Biermann K, Oosterhuis JW. Int J Androl. 2011 May 12. o Gonadal Pathology and Tumor Risk in Relation to Clinical Characteristics in Patients with 45,X/46,XY Mosaicism. Cools M, Pleskacova J, Stoop H, Hoebeke P, Van Laecke E, Drop SL, Lebl J, Oosterhuis JW, Looijenga LH, Wolffenbuttel KP; on behalf of the Mosaicism Collaborative Group. J Clin Endocrinol Metab. 2011 Apr 20. o Histopathological and molecular features of late relapses in non-seminomas. Mayer F, Wermann H, Albers P, Stoop H, Gillis AJ, Hartmann JT, Bokemeyer CC, Oosterhuis JW, Looijenga LH, Honecker F. BJU Int. 2010 Oct 18. o Histopathological and molecular features of late relapses in non-seminomas. Mayer F, Wermann H, Albers P, Stoop H, Gillis AJ, Hartmann JT, Bokemeyer CC, Oosterhuis JW, Looijenga LH, Honecker F. BJU Int. 2011 Mar;107(6):936-43. o Global DNA methylation in fetal human germ cells and germ cell tumours: association with differentiation and cisplatin resistance. Wermann H, Stoop H, Gillis AJ, Honecker F, van Gurp RJ, Ammerpohl O, Richter J, Oosterhuis JW, Bokemeyer C, Looijenga LH. J Pathol. 2010 Aug;221(4):433-42. o Gonadal tumours and DSD. Looijenga LH, Hersmus R, de Leeuw BH, Stoop H, Cools M, Oosterhuis JW, Drop SL, Wolffenbuttel KP. Best Pract Res Clin Endocrinol Metab. 2010 Apr;24(2):291-310. Review. o Seladin-1 and testicular germ cell tumours: new insights into cisplatin responsiveness. Nuti F, Luciani P, Marinari E, Erdei E, Bak M, Deledda C, Rosati F, Mazzinghi B, Danza G, Stoop H, Looijenga LH, Peri A, Serio M, Krausz C. o J Pathol. 2009 Dec;219(4):491-500Nuti F, Luciani P, Marinari E, Erdei E, Bak M, Deledda C, Rosati F, Mazzinghi B, Danza G, Stoop H, Looijenga LH, Peri A, Serio M, Krausz C. J Pathol. 2009 Dec;219(4):491-500. o A patient with an alpha-foetoprotein producing tumour. van Roon AH, ter Borg PC, Zondervan PE, Stoop H, de Man RA. Ned Tijdschr Geneeskd. 2009;153:A364. Dutch. o A novel SRY missense mutation affecting nuclear import in a 46,XY female patient with bilateral gonadoblastoma. Hersmus R, de Leeuw BH, Stoop H, Bernard P, van Doorn HC, Brüggenwirth HT, Drop SL, Oosterhuis JW, Harley VR, Looijenga LH. Eur J Hum Genet. 2009 Dec;17(12):1642-9. o Microsatellite instability, mismatch repair deficiency, and BRAF mutation in treatment- resistant germ cell tumors. Honecker F, Wermann H, Mayer F, Gillis AJ, Stoop H, van Gurp RJ, Oechsle K, Steyerberg E, Hartmann JT, Dinjens WN, Oosterhuis JW, Bokemeyer C, Looijenga LH. J Clin Oncol. 2009 May 1;27(13):2129-36. Epub 2009 Mar 16. o Evaluation of testicular biopsies for carcinoma in situ: immunohistochemistry is mandatory. van Casteren NJ, de Jong J, Stoop H, Steyerberg EW, de Bekker-Grob EW, Dohle GR, Oosterhuis JW, Looijenga LH. Int J Androl. 2009 Dec;32(6):666-74.

214

o Stem cell factor as a novel diagnostic marker for early malignant germ cells. Stoop H, Honecker F, van de Geijn GJ, Gillis AJ, Cools MC, de Boer M, Bokemeyer C, Wolffenbuttel KP, Drop SL, de Krijger RR, Dennis N, Summersgill B, McIntyre A, Shipley J, Oosterhuis JW, Looijenga LH. J Pathol. 2008 Sep;216(1):43-54. o FOXL2 and SOX9 as parameters of female and male gonadal differentiation in patients with various forms of disorders of sex development (DSD). Hersmus R, Kalfa N, de Leeuw B, Stoop H, Oosterhuis JW, de Krijger R, Wolffenbuttel KP, Drop SL, Veitia RA, Fellous M, Jaubert F, Looijenga LH. J Pathol. 2008 May;215(1):31-8. o Differential expression of SOX17 and SOX2 in germ cells and stem cells has biological and clinical implications. de Jong J, Stoop H, Gillis AJ, van Gurp RJ, van de Geijn GJ, Boer M, Hersmus R, Saunders PT, Anderson RA, Oosterhuis JW, Looijenga LH. J Pathol. 2008 May;215(1):21-30. o Gonadoblastoma-associated mixed germ cell tumour in 46,XY complete gonadal dysgenesis (Swyer syndrome): analysis of Y chromosomal genotype and OCT3/4 and TSPY expression profile. Ng SB, Yong MH, Knight LA, Lee VK, Nadarajah S, Stoop H, Looijenga LH. Histopathology. 2008 Apr;52(5):644-6. o Characteristics of testicular dysgenesis syndrome and decreased expression of SRY and SOX9 in Frasier syndrome. Schumacher V, Gueler B, Looijenga LH, Becker JU, Amann K, Engers R, Dotsch J, Stoop H, Schulz W, Royer-Pokora B. Mol Reprod Dev. 2008 Sep;75(9):1484-94. o A 40-year-old woman with a progressive periventricular white matter lesion. de Jong J, Stoop H, van den Bent M, Kros JM, Oosterhuis JW, Looijenga LH. Brain Pathol. 2008 Jan;18(1):103-4, 142. o Heterogeneous distribution of ITGCNU in an adult testis: consequences for biopsy-based diagnosis. van Casteren NJ, Boellaard WP, Dohle GR, Weber RF, Kuizinga MC, Stoop H, Oosterhuis WJ, Looijenga LH. Int J Surg Pathol. 2008 Jan;16(1):21-4. o Hepatoid adenocarcinoma of the gallbladder: a mimicker of hepatocellular carcinoma. van den Bos IC, Hussain SM, Dwarkasing RS, Stoop H, Zondervan PE, Krestin GP, de Man RA. Br J Radiol. 2007 Dec;80(960):e317-20. o Further characterization of the first seminoma cell line TCam-2. de Jong J, Stoop H, Gillis AJ, Hersmus R, van Gurp RJ, van de Geijn GJ, van Drunen E, Beverloo HB, Schneider DT, Sherlock JK, Baeten J, Kitazawa S, van Zoelen EJ, van Roozendaal K, Oosterhuis JW, Looijenga LH. Genes Chromosomes Cancer. 2008 Mar;47(3):185-96. o Noninvasive detection of testicular carcinoma in situ in semen using OCT3/4. van Casteren NJ, Stoop H, Dohle GR, de Wit R, Oosterhuis JW, Looijenga LH. Eur Urol. 2008 Jul;54(1):153-8. o Chromosomes and expression in human testicular germ-cell tumors: insight into their cell of origin and pathogenesis. Looijenga LH, Gillis AJ, Stoop HJ, Hersmus R, Oosterhuis JW. Ann N Y Acad Sci. 2007 Dec;1120:187-214. Epub 2007 Oct 2. Review. o High-throughput microRNAome analysis in human germ cell tumours. Gillis AJ, Stoop HJ, Hersmus R, Oosterhuis JW, Sun Y, Chen C, Guenther S, Sherlock J, Veltman I, Baeten J, van der Spek PJ, de Alarcon P, Looijenga LH. J Pathol. 2007 Nov;213(3):319-28. o JKT-1 is not a human seminoma cell line. de Jong J, Stoop H, Gillis AJ, van Gurp RJ, van Drunen E, Beverloo HB, Lau YF, Schneider DT, Sherlock JK, Baeten J, Hatakeyama S, Ohyama C, Oosterhuis JW, Looijenga LH. Int J Androl. 2007 Aug;30(4):350-65. o Why human extragonadal germ cell tumours occur in the midline of the body: old concepts, new perspectives. Oosterhuis JW, Stoop H, Honecker F, Looijenga LH. Int J Androl. 2007 Aug;30(4):256-63; discussion 263-4. o A genetic screen implicates miRNA-372 and miRNA-373 as oncogenes in testicular germ cell tumors. Voorhoeve PM, le Sage C, Schrier M, Gillis AJ, Stoop H, Nagel R, Liu YP, van Duijse J, Drost J, Griekspoor A, Zlotorynski E, Yabuta N, De Vita G, Nojima H, Looijenga LH, Agami R. Adv Exp Med Biol. 2007;604:17-46.

215 o Relevance of microRNAs in normal and malignant development, including human testicular germ cell tumours. Looijenga LH, Gillis AJ, Stoop H, Hersmus R, Oosterhuis JW. Int J Androl. 2007 Aug;30(4):304-14; discussion 314-5. Epub 2007 Jun 15. o Genomic and expression profiling of human spermatocytic seminomas: pathogenetic implications. Looijenga LH, Stoop H, Hersmus R, Gillis AJ, Wolter Oosterhuis J. Int J Androl. 2007 Aug;30(4):328-35; discussion 335-6. o Impact of the Y-containing cell line on histological differentiation patterns in dysgenetic gonads. Cools M, Boter M, van Gurp R, Stoop H, Poddighe P, Lau YF, Drop SL, Wolffenbuttel KP, Looijenga LH. Clin Endocrinol (Oxf). 2007 Aug;67(2):184-92. Epub 2007 Jun 4. o Gonadoblastoma arising in undifferentiated gonadal tissue within dysgenetic gonads. Cools M, Stoop H, Kersemaekers AM, Drop SL, Wolffenbuttel KP, Bourguignon JP, Slowikowska-Hilczer J, Kula K, Faradz SM, Oosterhuis JW, Looijenga LH. J Clin Endocrinol Metab. 2006 Jun;91(6):2404-13. o A genetic screen implicates miRNA-372 and miRNA-373 as oncogenes in testicular germ cell tumors. Voorhoeve PM, le Sage C, Schrier M, Gillis AJ, Stoop H, Nagel R, Liu YP, van Duijse J, Drost J, Griekspoor A, Zlotorynski E, Yabuta N, De Vita G, Nojima H, Looijenga LH, Agami R. Cell. 2006 Mar 24;124(6):1169-81. o The PTEN gene in locally progressive prostate cancer is preferentially inactivated by biallelic gene deletion. Verhagen PC, van Duijn PW, Hermans KG, Looijenga LH, van Gurp RJ, Stoop H, van der Kwast TH, Trapman J. J Pathol. 2006 Apr;208(5):699-707. o Genomic and expression profiling of human spermatocytic seminomas: primary spermatocyte as tumorigenic precursor and DMRT1 as candidate chromosome 9 gene. Looijenga LH, Hersmus R, Gillis AJ, Pfundt R, Stoop HJ, van Gurp RJ, Veltman J, Beverloo HB, van Drunen E, van Kessel AG, Pera RR, Schneider DT, Summersgill B, Shipley J, McIntyre A, van der Spek P, Schoenmakers E, Oosterhuis JW. Cancer Res. 2006 Jan 1;66(1):290-302. o Maturation delay of germ cells in fetuses with trisomy 21 results in increased risk for the development of testicular germ cell tumors. Cools M, Honecker F, Stoop H, Veltman JD, de Krijger RR, Steyerberg E, Wolffenbuttel KP, Bokemeyer C, Lau YF, Drop SL, Looijenga LH. Hum Pathol. 2006 Jan;37(1):101-11. o Germ cell lineage differentiation in non-seminomatous germ cell tumours. Honecker F, Stoop H, Mayer F, Bokemeyer C, Castrillon DH, Lau YF, Looijenga LH, Oosterhuis JW. J Pathol. 2006 Feb;208(3):395-400. o Sequence analysis of the protein kinase gene family in human testicular germ-cell tumors of adolescents and adults. Bignell G, Smith R, Hunter C, Stephens P, Davies H, Greenman C, Teague J, Butler A, Edkins S, Stevens C, O'Meara S, Parker A, Avis T, Barthorpe S, Brackenbury L, Buck G, Clements J, Cole J, Dicks E, Edwards K, Forbes S, Gorton M, Gray K, Halliday K, Harrison R, Hills K, Hinton J, Jones D, Kosmidou V, Laman R, Lugg R, Menzies A, Perry J, Petty R, Raine K, Shepherd R, Small A, Solomon H, Stephens Y, Tofts C, Varian J, Webb A, West S, Widaa S, Yates A, Gillis AJ, Stoop HJ, van Gurp RJ, Oosterhuis JW, Looijenga LH, Futreal PA, Wooster R, Stratton MR. Genes Chromosomes Cancer. 2006 Jan;45(1):42-6. o Identification of germ cells at risk for neoplastic transformation in gonadoblastoma: an immunohistochemical study for OCT3/4 and TSPY. Kersemaekers AM, Honecker F, Stoop H, Cools M, Molier M, Wolffenbuttel K, Bokemeyer C, Li Y, Lau YF, Oosterhuis JW, Looijenga LH. Hum Pathol. 2005 May;36(5):512-21. o Identification of recurrent chromosomal aberrations in germ cell tumors of neonates and infants using genomewide array-based comparative genomic hybridization. Veltman I, Veltman J, Janssen I, Hulsbergen-van de Kaa C, Oosterhuis W, Schneider D, Stoop H, Gillis A, Zahn S, Looijenga L, Göbel U, van Kessel AG. Genes Chromosomes Cancer. 2005 Aug;43(4):367-76. o Diagnostic value of OCT3/4 for pre-invasive and invasive testicular germ cell tumours. de Jong J, Stoop H, Dohle GR, Bangma CH, Kliffen M, van Esser JW, van den Bent M, Kros JM, Oosterhuis JW, Looijenga LH. J Pathol. 2005 Jun;206(2):242-9.

216

o Differentiation and development of human female germ cells during prenatal gonadogenesis: an immunohistochemical study. Stoop H, Honecker F, Cools M, de Krijger R, Bokemeyer C, Looijenga LH. Hum Reprod. 2005 Jun;20(6):1466-76. Epub 2005 Feb 25. o Rectal adenocarcinoma with choriocarcinomatous differentiation: clinical and genetic aspects. Verbeek W, Schulten HJ, Sperling M, Tiesmeier J, Stoop H, Dinjens W, Looijenga L, Wörmann B, Füzesi L, Donhuijsen K. Hum Pathol. 2004 Nov;35(11):1427-30. o Involvement of E-cadherin and beta-catenin in germ cell tumours and in normal male fetal germ cell development. Honecker F, Kersemaekers AM, Molier M, Van Weeren PC, Stoop H, De Krijger RR, Wolffenbuttel KP, Oosterhuis W, Bokemeyer C, Looijenga LH. J Pathol. 2004 Oct;204(2):167-74. o Resistance to platinum-containing chemotherapy in testicular germ cell tumors is associated with downregulation of the protein kinase SRPK1. Schenk PW, Stoop H, Bokemeyer C, Mayer F, Stoter G, Oosterhuis JW, Wiemer E, Looijenga LH, Nooter K. Neoplasia. 2004 Jul-Aug;6(4):297-301. o Pathobiological implications of the expression of markers of testicular carcinoma in situ by fetal germ cells. Honecker F, Stoop H, de Krijger RR, Chris Lau YF, Bokemeyer C, Looijenga LH. J Pathol. 2004 Jul;203(3):849-57. o Raman spectroscopic analysis identifies testicular microlithiasis as intratubular hydroxyapatite. De Jong BW, De Gouveia Brazao CA, Stoop H, Wolffenbuttel KP, Oosterhuis JW, Puppels GJ, Weber RF, Looijenga LH, Kok DJ. J Urol. 2004 Jan;171(1):92-6. o Stem cell factor receptor (c-KIT) codon 816 mutations predict development of bilateral testicular germ-cell tumors. Looijenga LH, de Leeuw H, van Oorschot M, van Gurp RJ, Stoop H, Gillis AJ, de Gouveia Brazao CA, Weber RF, Kirkels WJ, van Dijk T, von Lindern M, Valk P, Lajos G, Olah E, Nesland JM, Fosså SD, Oosterhuis JW. Cancer Res. 2003 Nov 15;63(22):7674- 8. o 12p-amplicon structure analysis in testicular germ cell tumors of adolescents and adults by array CGH. Zafarana G, Grygalewicz B, Gillis AJ, Vissers LE, van de Vliet W, van Gurp RJ, Stoop H, Debiec-Rychter M, Oosterhuis JW, van Kessel AG, Schoenmakers EF, Looijenga LH, Veltman JA. Oncogene. 2003 Oct 23;22(48):7695-701. o Xeroderma pigmentosum group a protein and chemotherapy resistance in human germ cell tumors. Honecker F, Mayer F, Stoop H, Oosterhuis JW, Koch S, Bokemeyer C, Looijenga LH. Lab Invest. 2003 Oct;83(10):1489-95. o Aneuploidy of human testicular germ cell tumors is associated with amplification of centrosomes. Mayer F, Stoop H, Sen S, Bokemeyer C, Oosterhuis JW, Looijenga LH. Oncogene. 2003 Jun 19;22(25):3859-66. o Morphology of testicular parenchyma adjacent to germ cell tumours. An interim report. Oosterhuis JW, Kersemaekers AM, Jacobsen GK, Timmer A, Steyerberg EW, Molier M, Van Weeren PC, Stoop H, Looijenga LH. APMIS. 2003 Jan;111(1):32-40; discussion 41-2. o POU5F1 (OCT3/4) identifies cells with pluripotent potential in human germ cell tumors. Looijenga LH, Stoop H, de Leeuw HP, de Gouveia Brazao CA, Gillis AJ, van Roozendaal KE, van Zoelen EJ, Weber RF, Wolffenbuttel KP, van Dekken H, Honecker F, Bokemeyer C, Perlman EJ, Schneider DT, Kononen J, Sauter G, Oosterhuis JW. Cancer Res. 2003 May 1;63(9):2244-50. o Molecular determinants of treatment response in human germ cell tumors. Mayer F, Stoop H, Scheffer GL, Scheper R, Oosterhuis JW, Looijenga LH, Bokemeyer C. Clin Cancer Res. 2003 Feb;9(2):767-73. o A novel case of infantile sacral teratoma and a constitutional t(12;15)(q13;q25) pat. Veltman I, van Asseldonk M, Schepens M, Stoop H, Looijenga L, Wouters C, Govaerts L, Suijkerbuijk R, van Kessel A. Cancer Genet Cytogenet. 2002 Jul 1;136(1):17-22. o Coamplification of DAD-R, SOX5, and EKI1 in human testicular seminomas, with specific overexpression of DAD-R, correlates with reduced levels of apoptosis and earlier clinical

217 manifestation. Zafarana G, Gillis AJ, van Gurp RJ, Olsson PG, Elstrodt F, Stoop H, Millán JL, Oosterhuis JW, Looijenga LH. Cancer Res. 2002 Mar 15;62(6):1822-31. o VASA is a specific marker for both normal and malignant human germ cells. Zeeman AM, Stoop H, Boter M, Gillis AJ, Castrillon DH, Oosterhuis JW, Looijenga LH. Lab Invest. 2002 Feb;82(2):159-66. o Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological implications. Stoop H, van Gurp R, de Krijger R, Geurts van Kessel A, Köberle B, Oosterhuis W, Looijenga L. Lab Invest. 2001 Jul;81(7):919-28. o Restricted 12p amplification and RAS mutation in human germ cell tumors of the adult testis. Roelofs H, Mostert MC, Pompe K, Zafarana G, van Oorschot M, van Gurp RJ, Gillis AJ, Stoop H, Beverloo B, Oosterhuis JW, Bokemeyer C, Looijenga LH. Am J Pathol. 2000 Oct;157(4):1155-66. o Comparative genomic and in situ hybridization of germ cell tumors of the infantile testis. Mostert M, Rosenberg C, Stoop H, Schuyer M, Timmer A, Oosterhuis W, Looijenga L. Lab Invest. 2000 Jul;80(7):1055-64. o Human papillomaviruses 6/11, 16/18 and 31/33/51 are not associated with squamous cell carcinoma of the urinary bladder. Westenend PJ, Stoop JA, Hendriks JG. BJU Int. 2001 Aug;88(3):198-201. o Identification of malignant cells in serous effusions using a panel of monoclonal antibodies Ber-EP4, MCA-b-12 and EMA. Stoop JA, Hendriks JG, Berends D. Cytopathology. 1992;3(5):297-302.

218 OCT3/4 POU5F1 SOX17 dAP EC